Combo MiniDV /SVHS JVC HR-DVS3 , VHS tape with tv recorded material
Settings 1-(B.E.S.T. on, Pic. Ct. Auto, Digital3R on, Digital TBC/NR on)
Settings2-(B.E.S.T. on, Pic. Ct. Edit, Digital3R off, Digital TBC/NR off )
LG HDD DVD Recorder RH199HS Hdmi 576i(PAL)YCbCr BT.601
Startech usb3hdcap / ATI HD 600 usb
Test patterns source
Dvd player Pioneer DV-868AVI output svideo
1 –JVC settings 1->svideo->LG->hdmi576i->Startech->Vdub->Huufyuv
2 – idem JVC deck settings 2
3 – JVC->svideo->AVT8710->ATi600->Vdub
4-JVC->DV-(firewire)-> Vdub/winDV (sdDV)
This particular combo as an input/output firewire that works also with the VHS player section .It also as directshow driver an can be used with virtualdub, also obviously with windv or scnenalyzer live.
Is the conventional workflow for this particular hardware still the best or the LG dvd recorder upscaling , filetring and ADC bring any benefit? Also the direct DV capture from a VHS source on the combo jvc deck brings significant loss of quality ( if quality loss is small or almost none edit could be done in a lossless format afterward )
samples for video 8 and also some detais about the LG DVD recorder can be seen in thread” DV conversion vs lossless capture” by vaporeon800
+ Reply to Thread
Results 1 to 30 of 116
I forgot to mention in the USB3HDCAP thread that you need to make sure the proc amp is set to default when capturing HDMI. The levels in your captures are only within 17-233. And you're sure it's set to output YCbCr?
There's a lot to digest here. My only other comment at the moment is that of the VHS captures, only the DV appears to be free of luma noise filtering (chroma is still frame-averaged, as seen at the scene change).
Last edited by vaporeon800; 14th Feb 2016 at 16:41.
Well no , I have used the sames values used for analog captures as the histogram was ok I have not changed them. I made a capture with the default values in the video proc. amp. and the captures now are within 16-240 and it seems better.I´m including the capoture equivalent to workflow 1 but with the default values on the startech pro.amp..Should I shrink the output color range to 16-235?
related to the output beeing YCbCr I have no control on the DVD recorder but the startech identifies the stream as BT601 signal colorimetry and as I think BT601 is indeed YCbCr , yes I think so.
Comparing 6 and 7 with each other it becomes obvious the AVT TBC adds a slight green cast, just like I was suspecting in the other thread recently.
Additionally it shows that it raises the black level severly and at the same time lowering the white level, it greatly reduces horizontal resolution and shifts the image a bit to the left.
In other words: I wouldn't ever recommend using the AVT as it degrades the image too much.
All caps except the DV one have spiky histograms, meaning there is a levels adjustment in only 8 Bits going on somewhere. This results in banding even though it is not visible to the naked eye here. Make sure you have the Proc-Amp at default settings (yes, even with the ATI600, as it seems) and disable all noise reduction settings in the LG (if any).
Last edited by Skiller; 15th Feb 2016 at 12:46.
The side effects of hardware filters (screenshots)
IMO it doesn't soften anywhere near that much in my NTSC samples. It could partly (or entirely?) be due to bad "alignment" between the capture window of the AVT's ADC and the DAC of the DVD player.
related to the output beeing YCbCr I have no control on the DVD recorder but the startech identifies the stream as BT601 signal colorimetry and as I think BT601 is indeed YCbCr , yes I think so.
Well I already had the idea that the tbc ( or at least this tbc) degrades somewhat the caps, however is the only way to avoid drop frames and the alike and for analog video is almost mandatory. The only other possibility is to use the dvd recorder in pass-through. However between a svideo-svideo pass-through only using filtering and tbc alike features vs also using the DVD recorder ADC it seems the last one gives better results. Also the startech seems to works better for digital hdmi inputs then for analog inputs.
All the first 7 caps have been made with Proc. Amp. settings different from default. With the startech over hdmi this is not a problem however with the ATI 600 usb I think that using default values I will get clipping and over saturation but I can try.
I have no noise reductions on the DVD recorder that I can access, only on the svhs deck .workflow 2 try to feed a raw stream into the dvd recorder ( all filtering an NR on the svhs deck off).
First capture 1 and the new capture 1 in the subsequent post show the difference between using different values and default values in the startech video proc. Amp.
Well I see the startech driver reports identical info with different video streams (YCbCr/RGB). However probably the hdmi stream is YCbCr as I have the dvd recorder set to record in MPEG-2 and MPGE-2 use YCbCr color space.
I don’t have to many settings for HDD/DVD recordings , picture quality XP,SP,LP,EP ,record aspect 16:9/4:3 , Record format MPEG-2 / MPEG4 , and multiple audio settings and other related to specific dvd recording features.
I don´t know if this can assure the hdmi stream is YCbCr or not but according to your post "The OUTPUT COLOR RANGE options are actually active with YCbCr input"
I tried to see if it´s active and the range changed when I select it from 16-240 to 30- 222.If this is also no proof, how is it possible to identify if it´s YCbCr or RGB?
Well but what I was really interested to known was if any of you having to your disposal this devices and workflows would choose as the best one and why or would you do any additional variant based on this devices
That post meant that it's active with YCbCr in addition to RGB (the option is normally used for RGB). I haven't thought of a way to identify whether the source is RGB, yet. My other capture devices dither RGB input when converting to 4:2:2, which makes it obvious, but the StarTech doesn't do that.
Sorry for sidetracking your thread. I'll share some further thoughts on the workflows once I have a chance to compare the files more.
v1=AviSource("7-DVD_svideo_ATI600.avi").Subtitle("none") v2=AviSource("6-DVD_svideo_AVT_ATI600.avi").ColorYUV(cont_y=36, off_y=-4, cont_u=-10, off_u=1, cont_v=-10, off_v=0).Crop(0,0,-2,0).AddBorders(2,0,0,0).ChromaShift(l=-1).Subtitle("AVT") Interleave(v1,v2) StackHorizontal(last, UtoY(), VtoY())
Last edited by jagabo; 15th Feb 2016 at 19:29.
Hmm, any reason why cont_y instead of gain_y?
They're basically the same thing except gain_y works around 0, cont_y around 128. Ie, cont_y is something like Y = ((Y - 128) * cont) + 128, whereas gain_y is like Y = ((Y - 0) * gain) + 0 (using unit values for easy parsing here). Since the bright level needed to go up and the dark level down cont_y was the more obvious choice. After the cont_y adjustment everything still needed to go down a little, hence the off_y adjustment. You can get the same results with gain_y and off_y using different values. And, of course, you can use the "smooth" versions to introduce less banding in the histograms.
Last edited by jagabo; 15th Feb 2016 at 21:34.
related to the LG YCbCr output it seems it can do both but probably the selection is made automatically depending of connected device.
Some info from the service manual
"AUDIO, VIDEO AND AUXILIARY DATA IS TRANSMITTED ACROSS THE THREE TMDS DATA CHANNELS.
THE VIDEO PIXEL CLOCK IS TRANSMITTED ON THE TMDS CLOCK CHANNEL AND IS USED BY THE
RECEIVER AS A FREQUENCY REFERENCE FOR DATA RECOVERY ON THE THREE TMDS DATA CHANNELS.
VIDEO DATA IS CARRIED AS A SERIES OF 24-BIT PIXELS ON THE THREE TMDS DATA CHANNELS.
TMDS ENCODING CONVERTS THE 8BIT PER CHANNEL INTO THE 10BIT DC–BALANCED.
VIDEO PIXEL RATES CAN RANGE FROM 25MHZ TO 165MHZ. THE VIDEO PIXELS CAN BE ENCODED IN
EITHER RGB,YCBCR 4:4:4 OR YCBCR 4:2:2 FORMATS. IN ALL THREE CASES, UP TP 24 BITS PER
PIXEL CAN BE TRANSFERRED."
Relies on LSI Logic's proprietary 'DVFX' video pre- and post-processing technologies
Last edited by FLP437; 15th Feb 2016 at 23:08.
No commentaries about the main thread subject?
I was really looking forward to have your expert advice's on this unorthodox workflow, they mean a lot to me and will provide the necessary data to make an informed and supported choice, as I need to decide what workflow to use.
What you think about the quality delivered by the alternative workflow ?
The conventional and the reference in this case to compare with will be
JVC SVHS/ Digital8 Canc.->svideo->AVT-8710(TBC)-ATI HD 600 usb->VirtualDub->Huufyuv
And the unconventional
JVC SVHS/Digital8 Canc.->svideo->DVD Recorder->Hdmi576i(PAL)->Startech usb3HDcap->VirtualDub->Huufyuv
What we are really trying to compare in terms of delivered results is the potential capacity of filtering and tbc alike functions provided by the dvd recorder (LSI DMN8653) versus the external full frame TBC AVT 8710 and the ADC capture quality from the ATI Hd 600 usb vs the DVD recorder internal ADC capture (Philips 7137d), also up-scaling and trans-code effectiveness eventually.
If this workflow makes sense for this particular set of equipment could any variant still improves it . Are the results better if the video source in not dnr´ed and line TBC filtered by the JVC deck for instance? Any other alternatives that can improve it ?
If you need any additional input data to give your opinion /advice, please let me know .
Looking forward for your opinions and advices
Some thoughts I have:
The built-in line TBC of the JVC does straighten lines nicely (removes jitter), while the AVT does not at all, nor does the LG or the ATI600.
Does the JVC allow to turn on it's TBC without also turning on Digital3R at the same time? If yes, I would turn OFF Digital3R and turn ON the TBC.
If both can only be turned on together I would turn both off and use a Panasonic DVD-recorder (ES10/EH52/ EH60) as a TBC that does remove jitter effectively and (unlike the JVC's TBC) does output a constant and stable signal. Or get one with HDMI output.
The LG does not seem to rival the Panasonic recorders in terms of TBC qualities, so if you want to go the HDMI route I would get a Pana with HDMI or only use the LG with the TBC of the JVC turned on. Same for the ATI600, either with the JVC's TBC turned on or with a Panasonic as a TBC between (replacing the AVT).
Thank you Skiller for your comments on the subject
Related to the JVC deck settings yes it does allow to turn off Digital3R, however NR and TBC are used together I must turn off or on both of then I cannot choose NR and TBC individually. Also video stabilizer (correct vertical vibration in the picture) is interconnected with this setting if I choose to set video stabilizer on TBC/NR goes automatically off and vice versa.
I have also two other inter related settings
Video calibration on/off - VCR checks the condition of the tape in use during playback and recording, and compensates to provide the highest-possible pictures.
AUTO: Provides the optimized picture benefits of the VIDEO CALIBRATION Picture System.
EDIT: Minimizes picture degradation during editing (recording and playback).
SOFT: Reduces image coarseness when viewing overplayed tapes containing a lot of noise.
SHARP: Clearer, sharper-edged picture when viewing images with lots of flat, same-colored surfaces such as cartoons.
I was already thinking to buy a used Panasonic DVD-recorder probably an EH65 that does have hdmi output ,exactly because I read it could be superior to the LG.
The problem using the ATI is it´s not enough to get the JVC with line tbc on or digital8 camcorder with TBC on only, I have a substantial number of inserted frames typical between 40-60 per hour or more. Only with external tbc or dvd record in pass-through I have a steady no drop/inserted frame record ,however it seems the tbc degrades somewhat the image so only the dvd recorder seems ok for the task.
From the samples have you an idea about picture grabber quality from the ATI vs the internal ADC included in the dvd recorder? depending on that I can capture on the ATI with the LG or Panasonic as pass-through only as TBC or capture on the dvd recorder itself LG /Panasonic and in this case outputting to HDMI. The startech is not very good to capture analog sources but I´m satisfied with hdmi transfers
I have yet to confirm but the quality of old analog tv recordings in the LG HDD recorder is quite high so it seems that eventually playback of effectively recorded stuff get better quality that in pass-through , it´s only a feeling not confirmed at this moment . I am able for example to up scale old stuff recorded to 720p with excellent quality.
I will have to test an osprey 210e borrowed from a friend but only in a 2-4 weeks time frame to see how it may compare with the ATI.
About the JVC's TBC and Digital3R being tied together, bummer, but, if you use a Panasonic either as a pass-through or capture with it through HDMI you won't need the internal TBC because the Pannys do remove jitter quite good. And about noise reduction... unless time is a big factor I would really do that in software.
I would recommend EDIT Picture Control on the JVC because you might gain a little bit of detail with this.
The histogram of the LG cap is less spiky (although I'm still wondering what is causing that, it must be either the LG or the Startech).
But if you decide to use an EH65 there is absolutely no point in using it in pass-through and capture with the ATI600 because all that'll do is add another conversion from digital to analog and back. Nothing to be gained by that. If you use the EH65 you're ought to capture via HDMI.
Last edited by Skiller; 17th Feb 2016 at 17:10.
Thank you for your additional info
I just ordered my new "used" Panny EH65 it will take however some time (2 weeks) to arrive.
Just to clarify as I didn't explain myself well.
The JVC deck does have separate controls for DigitalR3 (edge correction to the luminance signal to enhance details) and Digital TBC/NR ((Time Base Corrector) - removes jitter from fluctuating video signals to deliver a stable picture and Digital 3-DNR (Noise Reduction) which cuts noise and enables clear picture reproduction both are linked in this function.)
as so I can have Digital R3 off and DigitalTBC/NR on
Originally Posted by FLP437;2433235The JVC deck does have separate controls for [B
Well in this case no chance, is no
And that's why you'll want to use an external line TBC (DVD recorder pass through).
So my future workflows will be
JVC SVHS deck/ Digital8 Cam.->svideo-> Pana EH65->hdmi576i->startech usb3hdcap->Vdub-Huufyuv
With JVC deck settings
B.E.S.T. on, Pic. Ct. Edit, Digital3R off and Digital TBC/NR off )
And Digital8 Cam with TBC on , DNR off
I also have a JVC HR-S9500 butI think the HR-DVS3 is better
Eventually I will test the Pana only as a TBC in pass-through svideo-svideo with the ATI and the Osprey 210e. However knowing that I´m doing additional ADC /DAC and that quality can suffer with that.
I am relatively satisfied with the startech in terms of hdmi , I don’t know if any other device is significant better for hdmi transfer/record If blackmagic shuttle or 4k or other in the same price range external or internal (PCIe) are clear better I will be eventually willing to try or buy.
Last edited by FLP437; 17th Feb 2016 at 20:45.
There is only one matter that we have not really spoke about . The DV direct capture from the combo SVHS/MiniDV.
Is there significant loss of quality compared with the reference workflow or could it be an alternative with slight less quality good for some material with less quality requirements?
There are two additional points on which I would like to have your advice
VhS captures – I read in some posts that half D1 captures are superior to full D1 captures for subsequent MPEG2 encoding resulting in final DVD with the best quality.
Is this is a reality , how is this half D1 capture accomplished . capturing in 352x576 (PAL) format. I tried and at least one of the capture cards don’t accept this capture format. Find a card that supports this format ? Can it be done capturing in 720x576 and resizing to 352x576? Probably not . How exactly is this Hald D1 capture accomplished if indeed it produces better results?
I have found a used matrox MX02 Le with pcie adapter for about 350€ what I think is a reasonably/good price. Could this card bring any advantage to my workflow due to its considered good characteristics and also to the fact that is captures with a color depth of 10 bits . Probably not much if I am to record the hdmi stream, (is already digital can only be duplicated not improved I think) , eventually better results if I capture the analog signal but if I use the pana EH-65 I as pass-through I will be duplicating the A/D conversion . I don’t know if I will gain anything in terms of final result.
EDIT mode on the JVC should disable luma DNR regardless of the TBC/DNR button being switched on, thus allowing DNR "off" with line TBC. At least that seems to be the case with the NTSC samples I've seen. Chroma DNR is probably always-on, but the chroma artifacts aren't really noticeable on live action footage in playback IMO.
Only for comparison purposes I am including a sample with the following JVC settings
B.E.S.T. on , Pic. Control – edit , Digital3R-off, video stabilizer off , DigitalTBC/NR on
WF1 – JVC->svideo->LG-> hdmi576i->Startech->Vdub-> Huufyuv
WF2 – JVC-> svideo->LG->svideo->ATI600->Vdub->Huffyuv
I can ajust the Ati video proc. amp to avoid clipping in the histogram within virtualdub but even so, captures afterwards have luma from 20 to 245!....
Some thoughts about the Matrox MX02 LE w/ PCIe adapter I was thinking to eventually buy a used unit
First I know that most people would consider this like a” cannon to kill a mouse” due to the limited resolution of analog sources ( VHS,video8,Hi8 ) however if it can squeeze additional quality it can make sense to me, namely as the price for a used unit is high but not huge.
After reading a little about, I think this device could not work with virtualdub or if it does it could work with limitations or problems.
However it comes with Matrox A/V Tools application for capture/playback of audio and video, and is also full supported by Premiere Pro cs6.
With the a/v tools HD and SD analog, SDI, or HDMI can be captured into AVI files . Is possible to choose the uncompressed 8- or 10-bit codecs or the Matrox MPEG-2 I-frame codec at up to 300 Mb/s
It seems to support only video formats 720p and up but I´m not sure ,HDMI Y,Cb,Cr 4:2:2 (10-bit) / RGB (8-bit) .The specifications talk about SD analog but also that video formats support is from 720p and up ?
I am not sure if capturing in 10 bits color depth is really usefull, I dont know if Virtualdub support it and if it does support it after filtering and saving I probably get a final 8 bits color depth file but in the editing process I will have probably some gain.
Well I´m not sure if it could be used in this environment without problems and that I can benefit from its better specifications
I will wait to investigate a little more and have if possible some comments
My new (used) Panasonic DVD HDD recorder DMR-EH65 just arrived.
I made a initial capture using the next workflow
JVC HR-DVS3 (VHS)-> svideo->Pana DMR EH-65-> svideo->ATI 600usb->VDub->Huufyuv
It does have in picture settings a NR filter ( comb filter) the captures have been made with this comb filter off.
Is complicated to work with the ATI Proc. Amp to get luma in the captures between 16-235, it always tends to go near the maximum 255 hdmi are much easier to control
However I am unable to record through HDMI , I think some sort of HDCP problem . It cycles between no signal , HDCP, and attempts to handshake at different hdmi resolutions , but no effective output . It seems as there are some sort of HDCP problem in between the Panasonic and the startech usb3HDcap. I read elsewhere that it´s possible to overcome this problem inserting a HDMI splitter in between at least with a black magic 4k . I already ordered one, as soon as it arrives I will confirm that this problem can be solved.
I have used the scart av1 as source for the svideo signal as I have found in a German forum indication for preferential using this output in this dvd recorder and not the individual svideo, composite or component, they refer some sort of instability of video signals in these plugs. However the quality of my cables Scart->Svideo are good but not top quality ( profigold premium 1,5 m) and my svideo->svideo are top quality (QED QUnex 1m) .However the contact resistance in the scart plug must also be better ( smaller) as there is a large surface contact then in the svideo plug. In this case the better will be a top quality short scart- scart cable that is suggested in this site .I have an IXUS scart-scart 21 pins professional 0,6 m I will use it next time. However probably the eventually benefits are so small if any that are impossible to identify and the results will be exactly the same and all this discussion is only theoretical.
As soon as I have at my disposal the osprey 210e card I will do similar capture to see if the results are better them the ones from the ATI600 usb and as soon as the hdmi splitter arrives I will do also similar captures using HDMI if the HDCP problem is solved.
I am still thinking about buying the matrox MX02 Le but the fact that it seems only to accept HD Hdmi and accept sd analog but records it in HD (720p and up) if my interpretation of the specs are correct are forcing me to read more about it and get additional info to see the real interest of this capture device as it´s in the price limit i will wiling to pay.
I have some difficulties when it comes to evaluate a sample or a capture . Obviously as anybody else I can detect qualitative differences if they are evident , but when the samples are very similar in quality I have tremendous difficulties to identify what is the best capture and consequently identify the best workflow, settings whatever.
Are there any practical rules to identify defaults , problems and quality in general qualitatively speaking in whatvideo footage is concerned, Are there utilities , filters , plug-in that can quantify quality in any way through variables that are quality related , resolution , jitter , noise , etc that allow us to confirm that a qualitative feeling is really supported by quantitative measurements ?
I hope you're aware that only a small subset of HDMI splitters will strip HDCP. Most happily pass it along end-to-end, as they are supposed to.
I take it this latest capture is with the JVC's TBC on, since the head switch area is uncorrected.
Perhaps someone else can suggest worthwhile measurements for you.