speech recognition equipment/software to transcribe recordings

[p_s_92]

A group of people will be discussing a topic. The discussion needs to be
transcribed(converted to text so that it can be reviewed later).

I know Dragon Naturally Speaking software does that to some extent, but that needs to
be trained to recognize the speaker's voice(and our setting is a group of people discussing a topic) and after that also a human reviewer needs to listen to the recording to make corrections to the transcribed text.

Is there any other speech recognition software/equipment which can be used for this purpose?

I realize the accuracy of such a transcription might be only 70-75%

I know Google Voice does that for phone messages, but can that be used for this situation?

[freebird73717]

From what I know Dragon is top of the line for that kind of software. I've got a quadriplegic friend that has trouble typing and uses dragon. Yep it's not perfect and also you have to speak slow and enunciate your words for the program to work right.

I don't think you'll find anything that will work well for a group.

Honestly since you are going to be having someone listening to the recording to make corrections, why not just have that person just do a full transcribing job in the first place. Would probably be the easiest thing.

[p_s_92]

From what I know Dragon is top of the line for that kind of software. I've got a quadriplegic friend that has trouble typing and uses dragon. Yep it's not perfect and also you have to speak slow and enunciate your words for the program to work right.

I don't think you'll find anything that will work well for a group.

Yes the link explains it a little http://www.pcworld.com/article/138262/from_speech_to_text.html, so we may have to record each person's speech individually though how to do it I need to find out. Since it is a group discussion how can that be done? I mean person A will say something, then person B will talk, person A will then want to comment on what person B just said?
I have not done such recording before where different person's voice are on different tracks? What device would support recording on track 1 for person A's speech, track 2 for person B and then back to track 1 for person A? Sorry, if my question is naive, but I have not done this before. Which equipment would you recommend for doing such a recording? The office which we will be supporting for this project can provide a Sony Handycam DCR-DVD650 DVD 60X Digital Zoom http://accessories.us.dell.com/sna/products/Camcorders/productdetail.aspx?c=us&l=en&s=dhs&cs=19&sku=A2507 250&~ck=bncat&~bk=gr:bnMap:CategoryRec,g:NextStep, rk:&dgc=CJ&cid=24471&lid=566643&acd=10550055-1225267-u0t2171314f17fp0c0s441 if they can borrow it from somewhere.

Honestly since you are going to be having someone listening to the recording to make corrections, why not just have that person just do a full transcribing job in the first place. Would probably be the easiest thing.

Well, we don't have anyone who can type as fast as people would typically speak in a discussion. I know journalists use shorthand for such speeches and courtroom reporters can do that, but we don't have such people in our office.

Having a person doing it by listening to the recording later would mean he would have to spend a lot of time. Using a software/equipment would mean he hopefully would spend less time as it would only be for the corrections.

Thanks for your advice and time.

[freebird73717]

Is this going to be for a one time deal? If so then you might consider spending the money that you would spend on software and hire a professional transcriptionist.

Other than that I'm out of ideas. Maybe someone else will chime in later with something else. Good luck to you.

[neomaine]

My wife has been a transcriptionist for many years and though mostly within the health profession she's also worked with some courtrooms.

Everything that's been described here as a hinderance is spot on. What you need is a professional court transcriber who can handle the multiple conversations. They have the hardware and experience to handle what you need. They'll provide a full transcript for follow up.

And, to help things out, you'll need to have whoever the moderator for the meeting help out on making sure there aren't multiple people talking at the same time and setup ground rules.

I'd recommend against trying to setup multiple sets of hardware (once for each person) to record everything. Logistically, the recording will be easy. However, trying to figure out the timeline for all the people to put it back into a transcribed document would be a nightmare.

[Nelson37]

I would agree that doing this with software just is not going to work. Record the audio and have somebody re-type it, what you need is a foot-pedal start-stop switch, like a Dictaphone. The rate used to be a buck per page.

Adding up the time to setup the software and correct all the errors, the typist would be faster.

Dragon is good, but it is just not up to what you are trying to do.

[p_s_92]

Is this going to be for a one time deal? If so then you might consider spending the money that you would spend on software and hire a professional transcriptionist.

No, it is something which needs to be done about once a month which is why we are looking at the software option.

Other than that I'm out of ideas. Maybe someone else will chime in later with something else. Good luck to you.

Thanks for your advice and wishes.

[p_s_92]

My wife has been a transcriptionist for many years and though mostly within the health profession she's also worked with some courtrooms.

Everything that's been described here as a hinderance is spot on. What you need is a professional court transcriber who can handle the multiple conversations. They have the hardware and experience to handle what you need. They'll provide a full transcript for follow up.

We just cannot afford a professional court transcriber charges as this needs to be done once a month.

And, to help things out, you'll need to have whoever the moderator for the meeting help out on making sure there aren't multiple people talking at the same time and setup ground rules.

Yes, we will do that.

I'd recommend against trying to setup multiple sets of hardware (once for each person) to record everything. Logistically, the recording will be easy. However, trying to figure out the timeline for all the people to put it back into a transcribed document would be a nightmare.

For this we need to ask the participant say his name like Person A views before he starts speaking so that we can know when he started. Also, the moderator needs to note the sequence like Person A spoke, then Person C, then Person M and so on which could avoid some issues.

Thanks for your suggestions. I will keep them in mind.

[p_s_92]

I would agree that doing this with software just is not going to work. Record the audio and have somebody re-type it, what you need is a foot-pedal start-stop switch, like a Dictaphone. The rate used to be a buck per page.

Adding up the time to setup the software and correct all the errors, the typist would be faster.

Dragon is good, but it is just not up to what you are trying to do.

We just cannot afford a professional court transcriber charges as this needs to be done once a month. How much time would it take to setup software(do you mean training Dragon to recognize a person's voice)?

Also, http://www.franklin-square.com/transcription_per_line.htm shows some transcription costs but they seem more than a buck per page? Did you have a particular company in mind which does it for a buck per page or did you mean placing a project on odesk.com for overseas workers?

Thanks for your time.

[neomaine]

I would agree that doing this with software just is not going to work. Record the audio and have somebody re-type it, what you need is a foot-pedal start-stop switch, like a Dictaphone. The rate used to be a buck per page.

Adding up the time to setup the software and correct all the errors, the typist would be faster.

Dragon is good, but it is just not up to what you are trying to do.

We just cannot afford a professional court transcriber charges as this needs to be done once a month. How much time would it take to setup software(do you mean training Dragon to recognize a person's voice)?

Also, http://www.franklin-square.com/transcription_per_line.htm shows some transcription costs but they seem more than a buck per page? Did you have a particular company in mind which does it for a buck per page or did you mean placing a project on odesk.com for overseas workers?

Thanks for your time.

You won't really know a transcriber's cost until you find out. You may be pleasantly surprised.

You could also simply try them for the first time (and maybe a couple more). Its kind of low, but, once you get the idea of how its accomplished you may have a better idea of what you really need. Perhaps if it works out well the person in charge of budget will see that its the simplest way to go.

Don't forget the old addage: You can have fast, cheap or good. Pick any two.

[p_s_92]

You won't really know a transcriber's cost until you find out. You may be pleasantly surprised.

How much do you guess they would cost? I am located near Jacksonville in Florida. We asked a few college students who were asking 10$/hr for the task. For a 2 hour discussion, it would mean 10hours of labor or 100$/month. I guess professional transcriber if hired as a contractor would have much higher hourly rates if hired only for some amount of time?

You could also simply try them for the first time (and maybe a couple more). Its kind of low, but, once you get the idea of how its accomplished you may have a better idea of what you really need. Perhaps if it works out well the person in charge of budget will see that its the simplest way to go.
Don't forget the old addage: You can have fast, cheap or good. Pick any two.

http://www.franklin-square.com/transcription_comparison.htm shows per hour rate to be 25/hour(last line of page) so for 1 hour of audio it would mean 100$/hour if I am interpreting it correctly.

Are you suggesting we should try odesk.com so that it can be done overseas for 3-4/hour? Our office is not OK with sending such work overseas via a site like odesk.com.

What would you estimate the cost of transcribing a 1 hour video? I know it depends on how quickly the speaker's speak, the transcriber's expertise and so on, but what would you guess? Am I getting a proper ballpark figure from http://www.franklin-square.com/transcription_comparison.htm?

Thanks for your time.

[Cornucopia]

I would agree with the FAST vs. CHEAP vs. GOOD assessment.

My suggestion would be to get a multitrack recording software & card (I like ProTools, but there are a number of others).

Then, you mike each speaker individually (wired lavs), going into a pre-amp/mixer but not being mixed, just pre-amped and then direct out to the cap card.

Then, record live in multitrack, so each speaker has their own dedicated track. Use a clapper to start and stop the session, and have each speaker introduce themselves so you have correct channel assignment.

Using the multitrack software, after recording, clip/trim the recording down to where it's exactly the length bounded by the clapper's transient. That way all tracks will remain in sync even if they get split into individual files (probably likely).

Then, get a "transcribing" software, or some audio software that easily allows for smooth forward & reverse at multiple speeds (particularly 1/2 speed). The "transcribing" software often has facility for using a jog/shuttle wheel or pedal with which to make this job much easier and speedier.

Split the multiple channels into individual tracks that are supported by the transcribing software (or possibly in pairs).

Then, get multiple "students" or other reasonable labor (who at least have good knowledge of the language, and possibly of the particular jargon if it's scientific, etc).

The point here is you DON'T need true court/medical transcribers (and their rates) if you don't need to do it in real time (or near-real time). But you CAN split up the work and have it be done simultaneously (in parallel).

Hope that helps,

Scott

[p_s_92]

My suggestion would be to get a multitrack recording software & card (I like ProTools, but there are a number of others).

Then, you mike each speaker individually (wired lavs), going into a pre-amp/mixer but not being mixed, just pre-amped and then direct out to the cap card.

Would you recommend any particular multitrack recording card? I saw some at http://www.answers.com/topic/multi-track-recording-sound-cards but don't about the quality and what would be suited for our need?

Also, would Jokosher(http://en.wikipedia.org/wiki/Jokosher) or Audcaity(http://en.wikipedia.org/wiki/Audacity) do for a multi track recording software?

Then, record live in multitrack, so each speaker has their own dedicated track. Use a clapper to start and stop the session, and have each speaker introduce themselves so you have correct channel assignment.

Using the multitrack software, after recording, clip/trim the recording down to where it's exactly the length bounded by the clapper's transient. That way all tracks will remain in sync even if they get split into individual files (probably likely).

By clapper you mean a device, or something else? Sorry, if my question is naive, but I am not done any such recording.

Then, get a "transcribing" software, or some audio software that easily allows for smooth forward & reverse at multiple speeds (particularly 1/2 speed). The "transcribing" software often has facility for using a jog/shuttle wheel or pedal with which to make this job much easier and speedier.

Would Dragon be OK for transcribing and for moving forward/reverse at multiple speeds or do we need Dictaphone tools?

Split the multiple channels into individual tracks that are supported by the transcribing software (or possibly in pairs).

Can the splitting be done using the multi track recording software(Ardour or Jokosher or ProTools if we can get it)?

Hope that helps,

Scott

Yes, it certainly does. Thanks a lot. I highly appreciate the advice.

[Bjs]

Always digging about ... remembering outlook express can handle multiple identities ... so can microsofts speech recognition technology by handling "multiple profiles".

Link ... and down to "training the speech recognition engine"

Transferring profiles:

1: Each person will need to take time to create their uniquely named speech recognition profile ... the long the better

2: Browse to the download page for the Windows Speech Recognition Profile Tool (see Resources). Click the "Continue" button to validate your Windows installation, then scroll down the page and click the "Download" button next to "WSRProfile.exe." Save the file to the desktop.Double-click the file "WSRProfile.exe" on the desktop

3: Double-click the file "WSRProfile.exe" on the desktop

4: Click "Backup my speech profile"

5: Click "Default Speech Profile," or the name of the speech recognition profile that you want to back up. Click "Next."

6: Click the "Browse" button, and navigate to the removable storage device that you would like to use to move the profile to the new computer. Type a name such as "Speech Profile," then click "Save." Windows saves your profile. Click "Close."

7: Move the backed-up speech recognition profile to the second computer using a removable storage device like a CD or USB drive. Move the program "WSRProfile.exe" to the second computer as well, or download it again.

8: Run "WSRProfile.exe" on the second computer, and click "Restore my speech profile".

9: Click the "Browse" button. Navigate to the removable storage device you used to move the speech recognition profile to the second computer, and double-click the profile.

10: Click "Use the current speech profile" to make the backed-up speech recognition profile the default on the second computer.

11: Click "Close"

12: Repeat as required.

====

Maybe dragon naturally speaking has profiles as well but I use philips free speech (not free) ... in case anyone asks

[handyguy]

MY friend does CART, this is the court thing but she can also make it a closed caption file. Costs under $70 per hour. CART is what it's called.

[p_s_92]

Always digging about ... remembering outlook express can handle multiple identities ... so can microsofts speech recognition technology by handling "multiple profiles".

Link ... and down to "training the speech recognition engine"

Transferring profiles:

1: Each person will need to take time to create their uniquely named speech recognition profile ... the long the better

2: Browse to the download page for the Windows Speech Recognition Profile Tool (see Resources). Click the "Continue" button to validate your Windows installation, then scroll down the page and click the "Download" button next to "WSRProfile.exe." Save the file to the desktop.Double-click the file "WSRProfile.exe" on the desktop

3: Double-click the file "WSRProfile.exe" on the desktop

4: Click "Backup my speech profile"

5: Click "Default Speech Profile," or the name of the speech recognition profile that you want to back up. Click "Next."

6: Click the "Browse" button, and navigate to the removable storage device that you would like to use to move the profile to the new computer. Type a name such as "Speech Profile," then click "Save." Windows saves your profile. Click "Close."

7: Move the backed-up speech recognition profile to the second computer using a removable storage device like a CD or USB drive. Move the program "WSRProfile.exe" to the second computer as well, or download it again.

8: Run "WSRProfile.exe" on the second computer, and click "Restore my speech profile".

9: Click the "Browse" button. Navigate to the removable storage device you used to move the speech recognition profile to the second computer, and double-click the profile.

10: Click "Use the current speech profile" to make the backed-up speech recognition profile the default on the second computer.

11: Click "Close"

12: Repeat as required.

====

Maybe dragon naturally speaking has profiles as well but I use philips free speech (not free) ... in case anyone asks

Thanks a lot. Hope Dragon software can use the trained profiles. Your advice is very useful.

[p_s_92]

MY friend does CART, this is the court thing but she can also make it a closed caption file. Costs under $70 per hour. CART is what it's called.

Do you know what CART software your friend uses?

I did some searching for it and saw http://shop.procat.com/category.sc?categoryId=10 which is not affordable.

Thanks for the suggestion.

[handyguy]

Eclipse, but she changed it to something new. Costs like $2000....plus you need the equipment, a steno machine. That's why they charge so much.

[p_s_92]

Eclipse, but she changed it to something new. Costs like $2000....plus you need the equipment, a steno machine. That's why they charge so much.

I wish there was a low cost option.

2000 for software alone + steno machine costing some 800 for the low end at http://www.stenosearch.com/buyersguidem.html
is too high.

Does your friend know of any low budget options which court reporting students might use? I mean can we get a old version of the software and refurbished(or second hand) steno machine for a low cost?

[Cornucopia]

Would you recommend any particular multitrack recording card? I saw some at http://www.answers.com/topic/multi-track-recording-sound-cards but don't about the quality and what would be suited for our need?

Also, would Jokosher(http://en.wikipedia.org/wiki/Jokosher) or Audcaity(http://en.wikipedia.org/wiki/Audacity) do for a multi track recording software?


By clapper you mean a device, or something else? Sorry, if my question is naive, but I am not done any such recording.


Would Dragon be OK for transcribing and for moving forward/reverse at multiple speeds or do we need Dictaphone tools?


Can the splitting be done using the multi track recording software(Ardour or Jokosher or ProTools if we can get it)?

1. NO. The cards mentioned there are multi-channel cards (like 5.1 or 7.1 surround - and they're only so-so anyway). They are NOT multitrack cards. I'm talking 4,6,8,12,16,32,48,64,96 etc. tracks. You can find them if you google "Multitrack Interface" or "digital audio workstation". Software that deals with them are: ProTools, Cubase, MOTU Dig. Performer, SAW, ntrack, Nuendo, Sadie. Others are available which DON'T rely on on a hardware interface (like what you mentioned), but in your circumstances, the hardware interface is what you're needing (so they can be input simultaneously in parallel), so you want software that FULLY supports that hardware (and vice-versa). You'd probably want an 8-channel/track interface for what your setup would be like.

2. Clapper, or Clap-stick, is what you see on stereotypical MOVIE sets where the P.A. snaps the 2 pieces of wood/plastic together and says, "Take 2...". It is designed for helping to keep things like this IN SYNC. You want to do it at the start of your recording, and at the end.

3. THAT Dragon (the voice recognition one) isn't meant for transcribing in this manner (although I guess it's possible), and it certainly isn't OPTIMIZED for this purpose. They have other solutions for Medical/Legal Transcribing, but THOSE aren't really the full voice recog. types either, and they're WAY $$$$, as was previously mentioned. I'm talking more about things like Inqscribe, Transana, NCH Digital Transcriber, and solutions from Olympus & Sony (often using their "Voice Recorder" devices, although not tied to them). These OFTEN come with the USB footpedal and/or jog/shuttle wheel. You can often google "subtitle" or "transcribing" software. Do the homework first.

4. Splitting using mutlitrack software - YES. All that I know of can do that.


Sorry, but a lot of those other solutions suggested would cost way too much (and for software or labor alone), whereas you might be able to get away with $500 - $1000, maybe $2000 for a full-blown system (assuming you already have the PC) of hardware and software(s) to do all the jobs from start to finish. And they'd be much more versatile for other situations and projects that might come down the road.


Scott

[p_s_92]

1. NO. The cards mentioned there are multi-channel cards (like 5.1 or 7.1 surround - and they're only so-so anyway). They are NOT multitrack cards. I'm talking 4,6,8,12,16,32,48,64,96 etc. tracks. You can find them if you google "Multitrack Interface" or "digital audio workstation".

1. Did you mean something like
http://www.amazon.com/Zoom-R16-Multitrack-Controller-Interface/dp/B0035GHYCS
or
http://www.tascam.com/products/us-1641.html I got this from the link at http://www.associatedcontent.com/article/1083976/how_to_turn_your_computer_into_a_multitrack.html Am I on the right track?

2. Do you have any recommendations for a multitrack card which is of good quality at a affordable price which would suit our project.

Software that deals with them are: ProTools, Cubase, MOTU Dig. Performer, SAW, ntrack, Nuendo, Sadie. Others are available which DON'T rely on on a hardware interface (like what you mentioned), but in your circumstances, the hardware interface is what you're needing (so they can be input simultaneously in parallel), so you want software that FULLY supports that hardware (and vice-versa). You'd probably want an 8-channel/track interface for what your setup would be like.

ntrack Studio 6 looks the option for us due to low cost of 64. I wish there was some free/open source tool available.

3. By 8-channel/track interface you mean a kit like http://www.soundhousemusic.org/multitrack-recording-interface/
Sorry, if my questions are naive, but I am completely new to this.

Clapper, or Clap-stick, is what you see on stereotypical MOVIE sets where the P.A. snaps the 2 pieces of wood/plastic together and says, "Take 2...". It is designed for helping to keep things like this IN SYNC. You want to do it at the start of your recording, and at the end.

4. So, I could use any mechanism to create a sound like a hard hand clap to do the work a clapper stick is doing to keep costs low?

THAT Dragon (the voice recognition one) isn't meant for transcribing in this manner (although I guess it's possible), and it certainly isn't OPTIMIZED for this purpose. They have other solutions for Medical/Legal Transcribing, but THOSE aren't really the full voice recog. types either, and they're WAY $$$$, as was previously mentioned. I'm talking more about things like Inqscribe, Transana, NCH Digital Transcriber, and solutions from Olympus & Sony (often using their "Voice Recorder" devices, although not tied to them). These OFTEN come with the USB footpedal and/or jog/shuttle wheel. You can often google "subtitle" or "transcribing" software. Do the homework first.

I thought Dragon for voice recognition because it could convert user's speech into text with 70% accuracy and then a person could make the changes for the 30% errors. So, person needs to spend less time as only correction would be needed for about 30% of text so labor costs would be low. Although, Dragon needs to be trained with a person's voice to get good accuracy.

5. You are suggesting Inqscribe, Transana as they have mechanisms where a person could listen to the audio and then type what was spoken, pause the audio using keystrokes and type some more and so on?

The link http://www.theselva.com/free-transcription-software/ shows a list of free transcription software(to keep costs low).

Sorry, but a lot of those other solutions suggested would cost way too much (and for software or labor alone), whereas you might be able to get away with $500 - $1000, maybe $2000 for a full-blown system (assuming you already have the PC) of hardware and software(s) to do all the jobs from start to finish. And they'd be much more versatile for other situations and projects that might come down the road.

Yes, other solutions are expensive. Multi track cards are around 300$, software for them like ntrack Studio 6 around 64, then use Transana for transcribing which is free. The only variable cost is labor of the person listening to the tape and doing the corrections of words incorrectly identified by transcribing software and using college students for that would mean around 10$/hour.

6. Did I miss anything which could be expensive?

We have a laptop for the recording. We would need the multitrack cards, software which can process sound received on it like ntrack Studio 6, then some free software like Transana for splitting and transcribing and finally using some college students to do the correction of the transcription.

7. Did I estimate properly or am I leaving out anything?

Thanks a lot for your advice and time. That is very useful. Now, I am hoping we can do this at the lowest possible cost with reasonable quality.

[Cornucopia]

Yeah, you're on the right track...(those look good for your needs)

Actually, that Zoom R16 sounds like a decent prosumer rig. Record to multiple SDHC cards with the Zoom R16 (no PC needed - although you can interface with USB2), pop the cards into your laptop, use CubseLite (provided w/ZoomR16), nTrack or maybe even Jokosha or Audacity like you mentioned (since they won't need to be the ones doing the actual multitrack recording live, as the hardware box would be doing that). Save mono files all cut to the exact same length (for easy syncing and Timecode reference).

I don't know if UNTRAINED DragonDictate or DragonNS would get you any farther ahead, but if it does, more power to you. Try on a 5 minute sample scenario just like you intend; if it's a good enough starting point, OK. If not, you'll know to discard it and just transcribe from scratch.

Yeah, hand clap might do. Or rim shot or other percussion, or even one of those little metal/plastic clakkers.

Transcribing: you got my meaning. The jog/shuttle wheel and/or footpedal (with the right software) allows for forward+reverse and fast/slow play.


You could probably get all this stuff and try it out in a practice session and figure out how well it works for you. If it didn't, you might still be able to return it (check ahead of time). If it did, you're prepared and ready to go...

Scott

[Cornucopia]

TO ADD A LITTLE MORE:

What you didn't mention in your last post was whether you already had mikes. You want GOOD quality WIRED lavalier mikes. Cardioid if you can get 'em (most lavs are Omni's). This is what will cost the most (if you're getting 4, 5, 6, etc) as Good ones are probably minimum $200 a piece. But for sound isolations and ease of transcription, it would be worth it.

Scott

[lordsmurf]

I never cease to be amazed at the amount of equipment and software a person (or organization) will buy, understanding not a damned thing about the topic or how to use it, simply in the quest to avoid hiring a skilled professional.

I ******* hate people like you.

Quit being a cheap bitch. Pay a professional for this service. Long-term, it's going to be a lower cost anyway, as well as spread out over time.

Sorry, I just think that needed to be said.

[bendixG15]

.................................................. .. Sorry, I just think that needed to be said.

When it comes to the unknown, we all go through a learning process which can be painful and expensive. But we muddle through the best we can.

So cut us mortals some slack ...........

[lordsmurf]

Learning something new is understandable.

Being stubborn like a child, and refusing to get assistance, is not.
"I don't need you, I can do it myself."
"Well, no honey, you can't. We all have things we can't do, or shouldn't do. This is one of them."

Spending tons of money, in an area you know NOTHING about, is 100% foolish.
Especially when you consider that the alternative is professional quality work.

No slack on this one.

[p_s_92]

I never cease to be amazed at the amount of equipment and software a person (or organization) will buy, understanding not a damned thing about the topic or how to use it, simply in the quest to avoid hiring a skilled professional.

Pay a professional for this service. Long-term, it's going to be a lower cost anyway, as well as spread out over time.

Sorry, I just think that needed to be said.

If instead of lambasting someone, you had taken the time to read the whole thread you would have realized, we CANNOT afford a professional.

Quit blaming without understanding the needs of the other person.

As Scott mentioned, you DON'T need a professional, if you don't need real or near real time transcription.

I ******* hate people like you.
Quit being a cheap bitch.

Using f* words is not decent in a forum. If you cannot help, just don't curse. I hate people who don't want to help, but freely curse.

[lordsmurf]

So several thousand dollars for a room worth of hardware and software -- that you have no idea how to operate -- is somehow less costly than paying an expert maybe $100 or so per month? That sounds like the same kind of economics used by Goldman Sachs and Enron.

I don't need conservative nannying on vocabulary. What I have found, however, is seemingly people fail to listen/read at times, unless you shake them up a bit by using "evil curse words" -- then they suddenly pay attention. I don't get it. It's actually something I'm studying for my next advanced degree.

And I think you'll find that I'm quite helpful, if you view any of my 25,000 or so posts on this site. In this case, being helpful comes in the way of telling you what NOT to do.

[p_s_92]

So several thousand dollars for a room worth of hardware and software -- that you have no idea how to operate -- is somehow less costly than paying an expert maybe $100 or so per month?

READ the thread again. Scott clearly told, I can get away with 500-1000 NOT several thousand dollars for a room worth of hardware/software.

That sounds like the same kind of economics used by Goldman Sachs.

If you cannot interpret what I wanted and what others posted how can you understand anything about Goldman Sachs.

Which expert charges 100$ or so for a 2 hour recording per month including his equipment? I live near Jacksonville, Florida where the rates are not that cheap.

I don't need conservative nannying on vocabulary.

You need a lot more than nannying if you don't know what is decent behavior on a forum. To think, your occupation is Public Relations.

What I have found, however, is seemingly people fail to listen/read at times, unless you shake them up a bit by using "evil curse words" -- then they suddenly pay attention. I don't get it. It's actually something I'm studying for my next advanced degree.

When trying for your advanced degree, learn to READ, curse less. You are in Public Relations occupation, So to build the image of your company, you will curse in the newspaper so that people know that XYZ who is the Public Relation person for a company used such words. You will get a lot of attention for yourself and your company in the wrong way.

You don't know how to read, behave properly in a forum though you are in the Public Relations domain, yet want to comment on Goldman Sachs ability that they have no idea how to operate.

And I think you'll find that I'm quite helpful, if you view any of my 25,000 or so posts on this site. In this case, being helpful comes in the way of telling you want NOT to do.

In this thread you clearly did not behave properly at all. Cursing is not a way to advise people at all. As a Public Relations professional you should know better. More so, when another poster is helping the OP on how to do it within the budget needs of OP.

In this case, being helpful comes in the way of telling you want NOT to do.

What not do to, NOT want not to do.

[p_s_92]

Yeah, you're on the right track...(those look good for your needs)
Actually, that Zoom R16 sounds like a decent prosumer rig.

Yes, we will try to get it. How is this DA 2496 http://proaudiomusic.com/software/da-2496_audio_interface.htm and
Tascam-DP004-Pocketstudio at http://pro-audio.musiciansfriend.com/product/Tascam-DP004-Pocketstudio-Portable-4Track...der?sku=250011

Just trying to get an idea of different decent models available at an affordable price so that we can choose whichever is available.

Record to multiple SDHC cards with the Zoom R16 (no PC needed - although you can interface with USB2), pop the cards into your laptop, use CubseLite (provided w/ZoomR16), nTrack or maybe even Jokosha or Audacity like you mentioned (since they won't need to be the ones doing the actual multitrack recording live, as the hardware box would be doing that). Save mono files all cut to the exact same length (for easy syncing and Timecode reference).

So, we won't need nTrack if Cubselite is provided through ZoomR16 and there are open source tools like Audacity?

I don't know if UNTRAINED DragonDictate or DragonNS would get you any farther ahead, but if it does, more power to you. Try on a 5 minute sample scenario just like you intend; if it's a good enough starting point, OK. If not, you'll know to discard it and just transcribe from scratch.

Yes, that is what I plan to do. Untrained Dragon will not do. I have to train it by asking the speakers to do a 5/10 min speech, ensure speakers speak slowly, clearly.

You could probably get all this stuff and try it out in a practice session and figure out how well it works for you. If it didn't, you might still be able to return it (check ahead of time). If it did, you're prepared and ready to go...

Yes, this is what I am planning to do a test run and see how it goes.

What you didn't mention in your last post was whether you already had mikes. You want GOOD quality WIRED lavalier mikes. Cardioid if you can get 'em (most lavs are Omni's). This is what will cost the most (if you're getting 4, 5, 6, etc) as Good ones are probably minimum $200 a piece. But for sound isolations and ease of transcription, it would be worth it.

We don't have mikes. We have just one digital voice recorder.

I notice some mikes at http://www.fullcompass.com/category/Lapel-Lavalier-Microphones.html from 23 to 400.

A Cardioid is around 175 at http://www.fullcompass.com/product/286730.html.

This might be a naive question, but can you please advise? Can we remove a mike from a speaker after he has finished talking and have him give it to another person who is about to talk? This is in case we cannot get as many mikes as many speakers might be there in the discussion. I know we need to separate speaker A's comments from speaker B's comments using some method.

Omni mikes would have low quality which would make it unsuitable for this purpose?

Thank you Scott! You have been a great help.