Indeed, I overlooked this method because DV only camcorders do not have HDMI out. The HV20 is an HDV camcorder that also can record and play DV. With HDMI, the video is converted into uncompressed format, which is lossless, then you either can keep it uncompressed, or re-compress to another format. Since the original video is compressed, the HDMI route is no better than capturing DV over Firewire or USB, but it is no worse either except for huge file size. It is a viable choice if you cannot capture DV as is. Thanks for pointing out to this option, I forgot about it.
Thank you so much, sir. i have learned a lot with your videos. I am using d8 handycam, but I don't own a Mac to transfer my tapes over firewire, so i'm planning to use s-video, deinterlace and upscale the video with virtualdub as you explained in another video. Do you think that this could allow me to approach the desirable quality? Thanks again for taking the time to make this videos
Thank you! Yes, S-video is almost indistinguishable from DV provided you have a decent A/D converter. OTOH, if you have a Windows desktop, installing a Firewire card will cost you just $15 or less.
With a USB port you can capture lossless DV video to computer with a certain few Panasonic models of camcorder: * PV-GS75 / NV-GS75 * PV-GS150 / NV-GS150 * PV-GS180 / NV-GS180 (verified working) * PV-GS230 / NV-GS230 (verified working) * PV-GS250 / NV-GS250 * PV-GS300 / NV-GS300 (verified working) * PV-GS320 / NV-GS320 (verified working) * PV-GS330 / NV-GS330 * PV-GS400 / NV-GS400 (verified working) * PV-GS500 / NV-GS500 (verified working) The listed models which I have not yet verified will probably do the job also.
A pleasure to see your news vids ! Love things about that video technology. I guess Firewire is the only best method to preserve the best quality from the tape. I have digitized once ( Dv-Avi- Pal. 720×576 . 25fps interlaced) then with Virtulal Dub : deinterlace, double frame rate (25 to 50 fps ) and encoding into mp4 1080p/4K. ( for preserving a good bitrate ) So i think it's a good method...
Thanks! Yes, Firewire is the standard way to transfer DV (and HDV) video to a computer or to another DV/HDV camcorder or deck. Apple invented it, but Sony was instrumental to make it a part of DV standard. Sometimes USB works too, I think Panasonic offered it for two or three years: 2005, 2006 and maybe 2007, dont' know about other brands. After you get it to a computer, you process it in the same way as digitized VHS or Hi8, that is, if you want it to look good on RUclips then you need to deinterlace it into 50 or 60 fps and upscale to at least 720 lines. Although some DV camcorders can shoot progressive, the JVC GR-DVL9000/9500 could shoot progressive at 25p or 30p (depending on region) in 1999! ruclips.net/video/bWtiW5cIdwQ/видео.html
Thank you for diving into this topic! And for doing 1 to 1 digital vs analog sampling comparisons (and handheld! Which is really what this medium is about. No tripods). Unfortunately, seems like there's no real clear knowledge on this technology. Heck, there's more knowledge on Super8 cameras than on miniDV camcorders. So, we may not know enough on how to utilize them nowadays. The GS500 is indeed the only camcorder I heard that does USB capture. Yet, personally, I'm looking for the more compact travel-friendly cameras. I use the Sony DCR-SX33 which captures straight to SD card so it's digital. Yet, I wonder how that footage would compare to a digital/analog capture of a minDV tape camcorder. Panasonic made me hate miniDV since they all fail eventually, yet, this medium might have a higher quality than those newer SD-card recording camcorders. And I only talk about SD cameras, not HD. They don't have the miniDV look & feel.
"The GS500 is indeed the only camcorder I heard that does USB capture." - I believe all these can do it: GS75, GS150, GS180, GS250, GS300, GS320, GS400, GS500, but I personally tested only the GS500 and I've read reports of someone else having success with the GS320. Caveat emptor :) "I wonder how ... the Sony DCR-SX33 ... footage would compare to a digital/analog capture of a minDV tape camcorder." - When you say "digital/analog capture", do you mean capturing digital video through an analog port? Analog capture does not ruin in, just a bit less sharp and a bit more noise, still very watchable. Like most DV camcorders, the SX33 has CCD sensor, so no skew or jello or flash-banding. The SX33 records in MPEG-2, not an issue if it is well implemented, you should not see noticeable differences. Early Long-GOP implementations showed degradation from keyframe to keyframe, the video would "pulse". "Panasonic made me hate miniDV since they all fail eventually" - my Panasonic MiniDV camcorders work so far. I saw Sony failing and very many JVC camcorders failing. "yet, this medium might have a higher quality than those newer SD-card recording camcorders." - Depends on implementation. DV and MPEG-2 were developed at about the same time, but in the 1990s DV was less power-hungry and was symmetrical, so the same chip could be used for encode and decode. In the 2000s MPEG-2 became viable. "And I only talk about SD cameras, not HD. They don't have the miniDV look & feel." - Most HD camcorders have CMOS sensors, but some have CCD. CMOS has rolling shutter, CCD has global shutter. Also, CCD produce a "ray of light" on bright light sources, a vertical bright line caused by overdriving CCD cells. How does this look to you? ruclips.net/video/chD5LvN8YeM/видео.html and ruclips.net/video/zeWQBMUipWI/видео.html Not the best quality, I recompressed it several times, also I deinterlaced to 30p not 60p. Shot with a 3CCD HD camcorder.
@@ConsumerDV Many thanks! 1. About Mpeg-2 - I can edit it straight in Adobe apps, yet, not in Davinci. I must convert it to MP4 [as I don't want large files for such relatively low-res footage]. And magically, the mp4 files look a bit better/ sharper than the original mpeg2. My BIGGEST problem, though, is that I can't find any way to "preserve the smooth motion look" of these CCD camcorders [smooth soap opera look] - when editing it on the computer. There are "fields" and there is "framerate". And it seems like they are connected. I was sure because of the interlaced fields- where only "even" or "odd" number of fields are being shown over 60fps- then it's possible to have all fields [=progressive] within 25fps, and keep that smooth motion and export a 25fps [or 30fps] file. Yet, it seems like those two fields NEED 50-60fps to be displayed, and if I render something as 25fps progressive- then 1/2 half of the fields are deleted- turning the camcorder footage into a choppy mess. If you know exactly to which format, and setting- to convert such camcorder footage into usable+Smooth footage for editing- then please lemme know :) 2. Regarding capturing footage digitally [firewire] vs analog- it's surprising to see the quality isn't far off. Maybe I just imagined these camcorders record shadows [or contrasty scences] with very mushy / blotchy details in the shadows. But, who knows? Maybe you're right and they will all look more or less the same no matter which way you capture them 3. Nice footage from 2010. Looks good. I especially like to see those times, where the world was more sane :)
Both DV and MPEG-2 can be interlaced and progressive. I don't know how exactly to do it in Premiere, I use Sony Vegas (now Magix Vegas Pro) on Windows. I can edit interlaced video on an interlaced timeline, but I can choose to render any way I want: as interlaced or as 30p or as 60p. There is a deinterlacing option, and when I choose "Deinterlace: interpolate" and render as 60p, each field is converted into a full frame, and I get fluid motion. Or I can edit in 60p timeline, the render will look pretty much the same. Anyway, my render settings do not directly depend on timeline settings. Maybe in Premire it is different. Lately, I prefer to deinterlace first, and then edit as progressive-scan. If you want to try this option, feel free to check out my "VHS to 60p" video, if you skip to 9-minute mark you can see how interlaced video can be converted into progressive, which each field being converted into a complete frame: ruclips.net/video/XzY1Vo1occc/видео.html . The original VirtualDub cannot ingest MPEG-2, but VirtualDub2 can. QTGMC gives better results than built-in VirtualDub deinterlacers, and it is multi-platform, so worth taking a look. Setup may be a pain, but when it works, it is great. Mushy/blotchy details are on VHS/Hi8 videos, with chroma resolution being 40 samples horizontally at best. DV is much better. Funny that I cannot suggest a DV video of mine with "live" look, as I prefer to shoot at 25p/30p :) Here is a recent one, shot at 25p and slowed down to 24p, a bit overexposed and too pinkish, still I think it looks pretty good: ruclips.net/video/OdbG9rUx_BQ/видео.html
@@ConsumerDV I guess I don't have full understanding of what De-interlacing does. Maybe it doubles the fields so the missing field [odd/or Even field]- is being duplicated and placed there? But, then you also STILL need for some reason to have the footage render @60fps instead of 25fps, and that's what confuses me. I don't remember but I think I'm de-interlacing in After Effects or Davinci via a click on some setting. Meaning I'm interpolating or just interpreting the footage in a certain way [no need to re-render]- and then I place this footage on the timeline and edit. The old software in your tutorial- is it still all valid in 2024? With all those menus? It's a bummer to use something outside Adobe/Davinci but maybe it's a must. You also said something about QTGMC which I'd never heard of...
I apologize about the frame rate confusion. Each interlaced frame combines two independent fields - two separate images - taken in different time, you cannot combine them together into one frame, you will get combing or ghosting, your video will become soft. So, you need to un-comb your video, and convert each field into a separate frame. Simply speaking, every line in each field must be doubled, and fields must be converted to full frames, which is why deinterlacing is also known as "line doubling". There are two families of frame rates: one is 25/50, another is 30/60. It seems that you are located in a "PAL" country. At your location, TV is either 25 fps interlaced or 50 fps progressive, 25×2 = 50. In "NTSC" countries television is either 30 fps interlaced or 60 fps progressive: 30×2 = 60. I tried explaining interlacing here: ruclips.net/video/siOkPQcrIUY/видео.html but I forgot to mention that fields are never displayed simultaneously. On a CRT TV, fields are displayed in succession, one after another, you NEVER see them together. Each field is a separate picture. See also: 100fps.com There is a whole other can of worms when progressive content like a movie is sent via an interlaced stream, in this case you CAN and should combine fields; how do you combine them depends on the pulldown pattern. For 25fps, usually you just treat both fields as one frame, easy. "The old software in your tutorial- is it still all valid in 2024? With all those menus?" - sure, it works, Windows only, no Mac. "You also said something about QTGMC which I'd never heard of..." - it is a deinterlacing script for AviSynth. AviSynth is a runtime engine to run scripts for video manipulation. It is multiplatform, can run on Mac, Windows and Linux. The simplest script just loads QTGMC library, loads your video, then runs QTGMC on it. TL;DR: if you want your 25 fps interlaced video to have the same look on RUclips as on TV, render into 50p, and hopefully your NLE will take care of deinterlacing. Also, RUclips needs at least 720 lines to turn 50p on.
What about using HDMI? Like using a Canon HV20 camera which has HDMI out.
Indeed, I overlooked this method because DV only camcorders do not have HDMI out. The HV20 is an HDV camcorder that also can record and play DV. With HDMI, the video is converted into uncompressed format, which is lossless, then you either can keep it uncompressed, or re-compress to another format. Since the original video is compressed, the HDMI route is no better than capturing DV over Firewire or USB, but it is no worse either except for huge file size. It is a viable choice if you cannot capture DV as is. Thanks for pointing out to this option, I forgot about it.
@@ConsumerDV Thanks for the info!
Thank you so much, sir. i have learned a lot with your videos. I am using d8 handycam, but I don't own a Mac to transfer my tapes over firewire, so i'm planning to use s-video, deinterlace and upscale the video with virtualdub as you explained in another video. Do you think that this could allow me to approach the desirable quality?
Thanks again for taking the time to make this videos
Thank you! Yes, S-video is almost indistinguishable from DV provided you have a decent A/D converter. OTOH, if you have a Windows desktop, installing a Firewire card will cost you just $15 or less.
With a USB port you can capture lossless DV video to computer with a certain few Panasonic models of camcorder:
* PV-GS75 / NV-GS75
* PV-GS150 / NV-GS150
* PV-GS180 / NV-GS180 (verified working)
* PV-GS230 / NV-GS230 (verified working)
* PV-GS250 / NV-GS250
* PV-GS300 / NV-GS300 (verified working)
* PV-GS320 / NV-GS320 (verified working)
* PV-GS330 / NV-GS330
* PV-GS400 / NV-GS400 (verified working)
* PV-GS500 / NV-GS500 (verified working)
The listed models which I have not yet verified will probably do the job also.
A pleasure to see your news vids ! Love things about that video technology. I guess Firewire is the only best method to preserve the best quality from the tape. I have digitized once ( Dv-Avi- Pal. 720×576 . 25fps interlaced) then with Virtulal Dub : deinterlace, double frame rate (25 to 50 fps ) and encoding into mp4 1080p/4K. ( for preserving a good bitrate ) So i think it's a good method...
Thanks! Yes, Firewire is the standard way to transfer DV (and HDV) video to a computer or to another DV/HDV camcorder or deck. Apple invented it, but Sony was instrumental to make it a part of DV standard. Sometimes USB works too, I think Panasonic offered it for two or three years: 2005, 2006 and maybe 2007, dont' know about other brands. After you get it to a computer, you process it in the same way as digitized VHS or Hi8, that is, if you want it to look good on RUclips then you need to deinterlace it into 50 or 60 fps and upscale to at least 720 lines. Although some DV camcorders can shoot progressive, the JVC GR-DVL9000/9500 could shoot progressive at 25p or 30p (depending on region) in 1999! ruclips.net/video/bWtiW5cIdwQ/видео.html
Thank you very much!
Thank you for diving into this topic!
And for doing 1 to 1 digital vs analog sampling comparisons (and handheld! Which is really what this medium is about. No tripods).
Unfortunately, seems like there's no real clear knowledge on this technology. Heck, there's more knowledge on Super8 cameras than on miniDV camcorders. So, we may not know enough on how to utilize them nowadays.
The GS500 is indeed the only camcorder I heard that does USB capture. Yet, personally, I'm looking for the more compact travel-friendly cameras. I use the Sony DCR-SX33 which captures straight to SD card so it's digital. Yet, I wonder how that footage would compare to a digital/analog capture of a minDV tape camcorder.
Panasonic made me hate miniDV since they all fail eventually, yet, this medium might have a higher quality than those newer SD-card recording camcorders.
And I only talk about SD cameras, not HD. They don't have the miniDV look & feel.
"The GS500 is indeed the only camcorder I heard that does USB capture." - I believe all these can do it: GS75, GS150, GS180, GS250, GS300, GS320, GS400, GS500, but I personally tested only the GS500 and I've read reports of someone else having success with the GS320. Caveat emptor :)
"I wonder how ... the Sony DCR-SX33 ... footage would compare to a digital/analog capture of a minDV tape camcorder." - When you say "digital/analog capture", do you mean capturing digital video through an analog port? Analog capture does not ruin in, just a bit less sharp and a bit more noise, still very watchable.
Like most DV camcorders, the SX33 has CCD sensor, so no skew or jello or flash-banding. The SX33 records in MPEG-2, not an issue if it is well implemented, you should not see noticeable differences. Early Long-GOP implementations showed degradation from keyframe to keyframe, the video would "pulse".
"Panasonic made me hate miniDV since they all fail eventually" - my Panasonic MiniDV camcorders work so far. I saw Sony failing and very many JVC camcorders failing.
"yet, this medium might have a higher quality than those newer SD-card recording camcorders." - Depends on implementation. DV and MPEG-2 were developed at about the same time, but in the 1990s DV was less power-hungry and was symmetrical, so the same chip could be used for encode and decode. In the 2000s MPEG-2 became viable.
"And I only talk about SD cameras, not HD. They don't have the miniDV look & feel." - Most HD camcorders have CMOS sensors, but some have CCD. CMOS has rolling shutter, CCD has global shutter. Also, CCD produce a "ray of light" on bright light sources, a vertical bright line caused by overdriving CCD cells. How does this look to you? ruclips.net/video/chD5LvN8YeM/видео.html and ruclips.net/video/zeWQBMUipWI/видео.html Not the best quality, I recompressed it several times, also I deinterlaced to 30p not 60p. Shot with a 3CCD HD camcorder.
@@ConsumerDV Many thanks!
1. About Mpeg-2 - I can edit it straight in Adobe apps, yet, not in Davinci. I must convert it to MP4 [as I don't want large files for such relatively low-res footage]. And magically, the mp4 files look a bit better/ sharper than the original mpeg2. My BIGGEST problem, though, is that I can't find any way to "preserve the smooth motion look" of these CCD camcorders [smooth soap opera look] - when editing it on the computer. There are "fields" and there is "framerate". And it seems like they are connected.
I was sure because of the interlaced fields- where only "even" or "odd" number of fields are being shown over 60fps- then it's possible to have all fields [=progressive] within 25fps, and keep that smooth motion and export a 25fps [or 30fps] file. Yet, it seems like those two fields NEED 50-60fps to be displayed, and if I render something as 25fps progressive- then 1/2 half of the fields are deleted- turning the camcorder footage into a choppy mess.
If you know exactly to which format, and setting- to convert such camcorder footage into usable+Smooth footage for editing- then please lemme know :)
2. Regarding capturing footage digitally [firewire] vs analog- it's surprising to see the quality isn't far off. Maybe I just imagined these camcorders record shadows [or contrasty scences] with very mushy / blotchy details in the shadows. But, who knows? Maybe you're right and they will all look more or less the same no matter which way you capture them
3. Nice footage from 2010. Looks good. I especially like to see those times, where the world was more sane :)
Both DV and MPEG-2 can be interlaced and progressive. I don't know how exactly to do it in Premiere, I use Sony Vegas (now Magix Vegas Pro) on Windows. I can edit interlaced video on an interlaced timeline, but I can choose to render any way I want: as interlaced or as 30p or as 60p. There is a deinterlacing option, and when I choose "Deinterlace: interpolate" and render as 60p, each field is converted into a full frame, and I get fluid motion. Or I can edit in 60p timeline, the render will look pretty much the same. Anyway, my render settings do not directly depend on timeline settings. Maybe in Premire it is different.
Lately, I prefer to deinterlace first, and then edit as progressive-scan. If you want to try this option, feel free to check out my "VHS to 60p" video, if you skip to 9-minute mark you can see how interlaced video can be converted into progressive, which each field being converted into a complete frame: ruclips.net/video/XzY1Vo1occc/видео.html . The original VirtualDub cannot ingest MPEG-2, but VirtualDub2 can. QTGMC gives better results than built-in VirtualDub deinterlacers, and it is multi-platform, so worth taking a look. Setup may be a pain, but when it works, it is great.
Mushy/blotchy details are on VHS/Hi8 videos, with chroma resolution being 40 samples horizontally at best. DV is much better. Funny that I cannot suggest a DV video of mine with "live" look, as I prefer to shoot at 25p/30p :) Here is a recent one, shot at 25p and slowed down to 24p, a bit overexposed and too pinkish, still I think it looks pretty good: ruclips.net/video/OdbG9rUx_BQ/видео.html
@@ConsumerDV I guess I don't have full understanding of what De-interlacing does. Maybe it doubles the fields so the missing field [odd/or Even field]- is being duplicated and placed there? But, then you also STILL need for some reason to have the footage render @60fps instead of 25fps, and that's what confuses me.
I don't remember but I think I'm de-interlacing in After Effects or Davinci via a click on some setting. Meaning I'm interpolating or just interpreting the footage in a certain way [no need to re-render]- and then I place this footage on the timeline and edit.
The old software in your tutorial- is it still all valid in 2024? With all those menus?
It's a bummer to use something outside Adobe/Davinci but maybe it's a must.
You also said something about QTGMC which I'd never heard of...
I apologize about the frame rate confusion.
Each interlaced frame combines two independent fields - two separate images - taken in different time, you cannot combine them together into one frame, you will get combing or ghosting, your video will become soft. So, you need to un-comb your video, and convert each field into a separate frame. Simply speaking, every line in each field must be doubled, and fields must be converted to full frames, which is why deinterlacing is also known as "line doubling".
There are two families of frame rates: one is 25/50, another is 30/60. It seems that you are located in a "PAL" country. At your location, TV is either 25 fps interlaced or 50 fps progressive, 25×2 = 50. In "NTSC" countries television is either 30 fps interlaced or 60 fps progressive: 30×2 = 60.
I tried explaining interlacing here: ruclips.net/video/siOkPQcrIUY/видео.html but I forgot to mention that fields are never displayed simultaneously. On a CRT TV, fields are displayed in succession, one after another, you NEVER see them together. Each field is a separate picture. See also: 100fps.com
There is a whole other can of worms when progressive content like a movie is sent via an interlaced stream, in this case you CAN and should combine fields; how do you combine them depends on the pulldown pattern. For 25fps, usually you just treat both fields as one frame, easy.
"The old software in your tutorial- is it still all valid in 2024? With all those menus?" - sure, it works, Windows only, no Mac.
"You also said something about QTGMC which I'd never heard of..." - it is a deinterlacing script for AviSynth. AviSynth is a runtime engine to run scripts for video manipulation. It is multiplatform, can run on Mac, Windows and Linux. The simplest script just loads QTGMC library, loads your video, then runs QTGMC on it.
TL;DR: if you want your 25 fps interlaced video to have the same look on RUclips as on TV, render into 50p, and hopefully your NLE will take care of deinterlacing. Also, RUclips needs at least 720 lines to turn 50p on.