DV Tape vs. SDHC for a HD camera

[Cazz]

Hi

I was thinking about buying a HD camera and what I have see some camera have mini DV and some have SDHC to save that I have recording.

For me it sound that SDHC is the best and easy way to recording and then move it to my computer

But I dont understand why some camera have mini DV for a HD camera?

I have work alot with DV camera before and I like it but is not so fun to move it to the computer in realtime so if I have
recording something that take 1 hour it going to take 1 hour to move it to my computer

Something else I wonder is that what kind of quality is ot on Mini DV, I mean in SDHC it is most MTS or something like that but
what does it save in a MiniDV??

[minidv2dvd]

HD on miniDV tape is HDV. canon hv20/30/40 cams are the most popular using that format.

[Cazz]

ohh ok but it is same quality, I mean HDV and AVCHD (What I have understand that MTS and AVCHD is the same)

But save on a SDHC is better, I mean I can recording more then 1 hour and very easy to move it to my computer then use Mini DV???

or it maybe wrong to say that??

[edDV]

Read through this thread then ask again here.

"AVCHD vs. HDV"
https://forum.videohelp.com/topic365818.html?highlight=xdcam

[Cazz]

wow edDV, you did write alot and very good information (almost to much for me to understand )

so what I can understand HDV is a little better then AVCHD

I dont want to lose any quality when I have it on my computer for edit
My computer have a ok performance, not any extrem fast computer but is ok so I dont think (and hope) I going to have
any problem to edit in Vegas 9.

I'm no pro and I dont going to work as a pro so I going to buy a consumer camrecorder
I was thinking about Canon HV30 that use HDV or Canon HF100 that use (AVCHD)
I can even think about buying a camrecorder that have a hard drives but I dont know if that is good.

For me I'm just like to have a good camrecorder that I can recording and transfer the I have recording to my computer without
lose any quality. But I also like to recording maybe more then one hour if I going to recording some theatre performance.

[Marvingj]

AVCHD actually performs somewhat better than HDV bit-for-bit, but everytime you reduce bitrate and maintain or improve perceived visual quality, you need more processing torque to decode it as the compression gets more and more sophisticated.

[Cazz]

ohh ok, thanks for that information.

But still have problem what I going to choose


What I can see

HDV:
+ Very good quality, maybe the best quality
- Not so easy to transfer from a camcorder to computer.
- Realtime transfer, so if I have recoding a hour it going to take a hour to transfer
- Only can recording a hour for high quality (Have not see other tape that can recording more then 60 min)

AVCHD
+ Very easy to transfer from a card to computer
+ Can recording more then a hour dependence on the size of the card.
- Maybe dont give the best quality??


This is what I have understand right now, you can tell if I have misunderstand something

[edDV]

wow edDV, you did write alot and very good information (almost to much for me to understand )

so what I can understand HDV is a little better then AVCHD

I dont want to lose any quality when I have it on my computer for edit
My computer have a ok performance, not any extrem fast computer but is ok so I dont think (and hope) I going to have
any problem to edit in Vegas 9

AVCHD will hold quality for cuts editing but if you intend to filter or add effects there will be greater processing loss than HDV. Also timeline performance is more sluggish for AVCHD as you try to search for an edit point or preview an effects edit. This is because the CPU alone is too slow for real time decode.

This can be improved by using a digital intermediate on import such as Cineform Neo Scene ($129) which also will help speed HDV.
http://www.cineform.com/neoscene/

I'm no pro and I dont going to work as a pro so I going to buy a consumer camrecorder
I was thinking about Canon HV30 that use HDV or Canon HF100 that use (AVCHD)
I can even think about buying a camrecorder that have a hard drives but I dont know if that is good.

For me I'm just like to have a good camrecorder that I can recording and transfer the I have recording to my computer without
lose any quality. But I also like to recording maybe more then one hour if I going to recording some theatre performance.

The HV30 has most the features of a Prosumer camcorder and will work well with external audio accessories such as import of house audio.

The one hour tape is an issue but I get around that by recording direct to a laptop over IEEE-1394. For long performance or conference proceedings I record to tape and laptop at the same time. When it is time to change tapes, this can be done with the laptop still recording so nothing is lost. The tapes plus the clip covering the tape change become the backup.

[edDV]

ohh ok, thanks for that information.

But still have problem what I going to choose


What I can see

HDV:
+ Very good quality, maybe the best quality
- Not so easy to transfer from a camcorder to computer.
- Realtime transfer, so if I have recoding a hour it going to take a hour to transfer
- Only can recording a hour for high quality (Have not see other tape that can recording more then 60 min)

AVCHD
+ Very easy to transfer from a card to computer
+ Can recording more then a hour dependence on the size of the card.
- Maybe dont give the best quality??


This is what I have understand right now, you can tell if I have misunderstand something

I'd add to the plus for the HV30 a community of advanced amateurs and pros that have solved most all issues and accessory configurations. There is even a full forum devoted to these models www.HV20.com.

The one hour issue is rare but I described a way to record through a tape change using a laptop. There are also external hard drive solutions (See Firestore FS4, FS5) if you will be doing much of this.
http://www.videoguys.com/Guide/RE/0x6fe9251a0226fe1478f00a47b0b1f9b7/Videoguys+FAQ+Foc...dd61e9342.aspx

Also figure media cost if you will be doing a large volume of work (e.g. $3/hr. for tape vs. ~$30/hr for flashram). Tape gives an instant backup. Flash requires transfer and backup strategies.

If you use the digital intermediate format to speed the actual edit session, there is still an ~1x import step that will be the same for HDV or AVCHD.

The additional pro for AVCHD is ability to separate a backup strategy from access to the camcorder. This can also be done for HDV but if you intend to use HDV tape for backup, you need to maintain access to an HDV camcorder.

[Cazz]

ohh ok
Well I did remember when I did recording a theatre performance with a DV camera with IEEE-1394 and a computer.
I did see that quality was not so good. Maybe was the laptop that did have poor firewire (I did use winDV)

But that you change the tape and recording at the same time was a good idea.

[edDV]

ohh ok
Well I did remember when I did recording a theatre performance with a DV camera with IEEE-1394 and a computer.
I did see that quality was not so good. Maybe was the laptop that did have poor firewire (I did use winDV)

But that you change the tape and recording at the same time was a good idea.

A recent laptop should be able to handle this. Older laptops may have problems.

Live Firewire recording to a computer works with either DV or HDV formats. For HDV, monitoring of the stream takes the most CPU load. You can record without monitoring with DVIO (DV) or HDVsplit (HDV).

[poisondeathray]

AVCHD will hold quality for cuts editing but if you intend to filter or add effects there will be greater processing loss than HDV.

What's the explanation or reasoning for "greater processing loss"?

Another possible storage option to look at is HD (hard drive) or hybrid storage camcorders (some have both HD + flash). This introduces more pros/cons.

[edDV]

AVCHD will hold quality for cuts editing but if you intend to filter or add effects there will be greater processing loss than HDV.

What's the explanation or reasoning for "greater processing loss"?

AVCHD uses MPeg4 h.264 which was designed as a final distribution format not an editing format. HDV is derived from Sony's XDCAM/IMX family of editable MPeg2. H.264 will suffer more recode loss. Beyond that most consumer edit programs recode twice for AVCHD to allow use of slower computers and to keep temp files resonably small (essentially using MPeg2 as an intermediate).

Typical consumer path
AVCHD --> recode to MPeg2 --> edit MPeg2* --> Export recode to AVCHD (h.264)

"Native AVCHD" (e.g. Vegas Pro)
AVCHD --> decompress on the fly for preview**, decompress to RGB for processing --> Smart render back to AVCHD

Digital Intermediate path
AVCHD --> recode to Cineform Digital Intermediate (all I frames) --> edit Cineform --> Export encode to AVCHD.


* effects or filtering requires another recode
MPeg2 --> RGB --> Mpeg2

** decompress on the fly requires a fast CPU and still seems sluggish

[poisondeathray]

I agree if you you use an MPEG2 intermediate for AVCHD, there will be more loss, but this is not the case I'm interested in.

Maybe I'm missing something? Are you saying those sections that are re-rendered have less loss with "editable MPEG2" compared to AVCHD? I don't see how this is possible in technical terms. These sections have to be re-rendered regardless of format.

I agree if there is a smart render function, and pass through for those untouched segments is functional (not re-rendered), that this will be better than re-rendering it.

[Cazz]

One question.

If I like to save source on my disk. insted of tape (If I now going to buy a camera that have HDV I going to save some on tape but most I going to have on disk)
What format can I have it so I dont loose quality but still easy to edit?
And if I can choose that dont take so much space.
I know if I like to save space I can't get so good quality but I like to know

[poisondeathray]

@Cazz - you can just keep it in original hdv format, it's usually some variant of transport stream like .m2t. Much easier to edit and perfect quality. If you compress it, you will lose quality; no way around this

[edDV]

I agree if you you use an MPEG2 intermediate for AVCHD, there will be more loss, but this is not the case I'm interested in.

Maybe I'm missing something? Are you saying those sections that are re-rendered have less loss with "editable MPEG2" compared to AVCHD? I don't see how this is possible in technical terms. These sections have to be re-rendered regardless of format.

I'm saying that MPeg2 suffers less recode loss than h.264 and that many workflows add additional recodes for AVCHD.

A typical HDV consumer workflow
HDV --> edit MPeg2 --> Export HDV
(Apple's iMovie also shrinks 1440x1080i to 960x540p for faster performance)

The big advantage for HDV is less CPU required for preview or conversely better timeline scrubbing for a fast CPU. The native HDV path is the same as native AVCHD but preview and timeline performance takes less CPU.

I agree if there is a smart render function, and pass through for those untouched segments is functional (not re-rendered), that this will be better than re-rendering it.

I'm not certain Vegas Pro is doing this for AVCHD but they do for DV and I'm fairly sure for HDV GOPs.

[Cazz]

poisondeathray:
ohh ok so when I use the capture on Vegas 9 it save it like .m2t or something similar format.

well I did could guess that, I mean when I did use capture on a DV camera it save it as AVI DV

Is always alot to think about when it time to invest in a new camrecorder.
I dont like to buy a camrecorder that is technical old after 5 years.

The Canon HV30 is a little expensive then Canon HF100 but that is ok if Canon HV30 is better.

I have found a file exampel of Canon HF100 in .MTS, is that possible to find a exempel about Canon HV30 that I can download and look at it and use it on Vegas to see what I like it or not?

[poisondeathray]

Cazz - I mean keep it the original captured format. You don't necessarily have to use vegas. e.g. many people use hdvsplit

edDV - I think I get what you are saying: More sections of using AVCHD as input will be re-encoded and thus lower quality compared to MPEG2, because of MPEG2 uses smart render and doesn't re-encode uncessary sections ?

Because "MPeg2 suffers less recode loss " implies something different to me, I thought you meant on sections that required re-encoding (due to effects, overlays , whatever) somehow MPEG2 was magically better than AVCHD, which is not the case.

[edDV]

In the Canon range, the HV30 has a 1/2.7" sensor vs. 1/4" in equivalent HF models. That gives advantage for low light. The HF-S100 has a similar size 1/2.6 sensor.

I haven't compared all the other camcorder features.

[edDV]

edDV - I think I get what you are saying: More sections of using AVCHD as input will be re-encoded and thus lower quality compared to MPEG2, because of MPEG2 uses smart render and doesn't re-encode uncessary sections ?

Because "MPeg2 suffers less recode loss " implies something different to me, I thought you meant on sections that required re-encoding (due to effects, overlays , whatever) somehow MPEG2 was magically better than AVCHD, which is not the case.

I'm saying that on a simple decode-encode h.264 suffers greater recode loss than MPeg2.

Beyond that HDV may or may not have a more efficient workflow depending on the product. Most edit products recode every frame. Some pro products pass unprocessed GOPs without recode. For high end editors this means tracking every GOP through filters, translations and compositing. I think they take short cuts (less smart rendering).

Note that "smart rendering" only works when the export format is identical in every way to the source format. Everything else gets recoded.

[Cazz]

poisondeathray:
well I can use that program instead Vegas when I transfer to the computer.

edDV:
Well that is good, I mean if I recording a theatre that have not so high light I like to have good quality. Or if I recording at a zoo and I'm inside some place dark.

Do anyone know if that is possible to find a exampel of a file that have been recording in Canon HV30. I know some I can see at www.vimeo.com but they are not original.

[edDV]

See individual camcorder reviews at www.camcorderinfo.com for low light comparisons. They have excellent reviews but not on all models.

[poisondeathray]

I'm saying that on a simple decode-encode h.264 suffers greater recode loss than MPeg2.

Assuming both your sources were of similar "quality", how is that possible? The software "sees" the exact same decoded frame. Maybe I'm completely misunderstanding what you mean by "recode loss"? You keep on repeating that, but maybe I'm too stoopid to understand.

e.g. If you have MPEG2 blu-ray and an h.264 blu-ray, and assuming they were of the same "quality", are you saying somehow re-encoding the h.264 blu-ray that it will suffer greater loss than the MPEG2 version? (assuming the same export settings) ?

If anything, I would say the export settings are more important than the import format, assuming "similar quality" for import. e.g. HDV=>h.264 would be less lossy than HDV=>MPEG2, Or AVCHD=>h.264 would be less lossy than AVCHD=>MPEG2 given similar bitrates

[edDV]

It will take more time than I have to explain but for a given source (frame size, frame rate, etc.), different codecs can be optimized for different purposes. HDV MPeg2 was optimized for better performance when decoded and recoded. h.264 was optimized as a one way distribution codec (final product). As such it can display high quality with one decode (in the player or display) but can't handle a decode, then recode without significant loss.

Broadcast video requires repeated recodes along the transmission path and for adding graphics overlays, etc. They use MPeg2 at higher bit rates to handle these recodes. h.264 would be a last step codec such as for satellite to home or to Blu-Ray disc. Tradeoff advantage for h.264 is lower bit rates (file size) while maintaining quality for display decode.

I'll be back later.

[Cazz]

wow thanks edDV, that is very much to read in www.camcorderinfo.com and I going to read all about it.

Just one question, I know about Tape and card, but what about a HDD?
What does it save, is that same as the AVCHD or is it something else??

[edDV]

wow thanks edDV, that is very much to read in www.camcorderinfo.com and I going to read all about it.

Just one question, I know about Tape and card, but what about a HDD?
What does it save, is that same as the AVCHD or is it something else??

Same as AVCHD.

[Cazz]

wow thanks edDV, that is very much to read in www.camcorderinfo.com and I going to read all about it.

Just one question, I know about Tape and card, but what about a HDD?
What does it save, is that same as the AVCHD or is it something else??

Same as AVCHD.

ok, because I did read the Canon HG-20 have a best recording in 24 Mbps (MXP+) and what I remember AVCHD on SDHC only have around 16 Mbps, but I can guess the is the codec of h264 that make it less good to edit then mpeg2.

The only funny is that I dont see if the HG-20 have a firewire

Did read that "Most of the AVCHD camcorders for 2008 are recording in 1920 x 1080, while HDV is forever locked at 1440 x 1080, then stretched for playback............but it’s only a matter of time before they crack the code" but I dont think I going to see any difference between 1920 x 1080 and 1440 x 1080

As you can see I maybe have very stupid question but have never work with this before.
I mean I have work with codec and edit but not so much about the technical about HD.

I maybe going to buy a HDV anyway but like to have so much information as possible before I buy a new camrecorder.

[Braindrain]

Just another data point, camcorderinfo stated the newest Canon AVCHD camcorders have a picture quality "near identical" (or something like that) to the HV30.

As for myself, I could not see a difference between HF20 and HV30 pre-edit vids. Yeah, there are some performance differences in lighting conditions, but that's a function of the camcorder itself and not the recording format.

edDV made a comment about coding to a lossless intermediate codec (Cineform) which is a very acceptable solution for video editing and subsequent encoding. Lastly, your encoder (and related specs) is just as important.

For me, it was a toss-up between the HV30 and the HF200. I went with the HF200 for various reasons, which include convenience (no more lugging multiple tapes and waiting forever for vid capture) without sacrificing quality.

I'd suggest you actually try the different formats and decide for yourself which works best as there are pros and cons to both.

[poisondeathray]

It will take more time than I have to explain but for a given source (frame size, frame rate, etc.), different codecs can be optimized for different purposes. HDV MPeg2 was optimized for better performance when decoded and recoded. h.264 was optimized as a one way distribution codec (final product). As such it can display high quality with one decode (in the player or display) but can't handle a decode, then recode without significant loss.

Broadcast video requires repeated recodes along the transmission path and for adding graphics overlays, etc. They use MPeg2 at higher bit rates to handle these recodes. h.264 would be a last step codec such as for satellite to home or to Blu-Ray disc. Tradeoff advantage for h.264 is lower bit rates (file size) while maintaining quality for display decode.

I'll be back later.

Thanks, looking forward to your technical explanation.

It seems a bit contradictory to me. Once you have the "high quality decode", the identical frame gets fed into the encoder. SO assuming it's the "same quality" input, it doesn't matter if it's lossless, MPEG2, Divx, h.264, or VP7 etc... e.g. if you took a lossless source, the only difference in output would be the export settings. Subsequent generations would be dependent on the export settings, of course higher quality upstream will result in higher quality downstream.

h.264 offers better compression than MPEG2, has clearly better objective measures (psnr, ssim) and subjective scores by a longshot (not even close here). If you were to make a plot, varying the bitrate on the x-axis, with psnr or ssim on the y-axis, MPEG2 never reaches the same quality at any bitrate, and never intersects the h.264 line at any point (h.264 eventually becomes lossless, MPEG2 does not). i.e. at any given bitrate, h.264 will have better quality, or obtains a certain quality level at a lower bitrate. Therefore, subsequent encodes using MPEG2 at the same bitrate will result in greater loss and noise / compression artifacts than h.264.

Hypothetical situation: Let's say you have a 1080p24 lossless source e.g. Lagarith encode of some CGI animation (or any content)

Scenario A: Encode that source @15Mbps, and then use that as input for subsesquent stage using MPEG2@15Mbps, and one more time (3 generations)

Scenario B: Same thing @15Mbps for h.264 (3 generations)

Which do you think will look better? both objectively (with psnr, ssim scores, difference masks & digital subtraction), or subjectively?