HVR-A1 or another PD-170

[Mike Duncan]

Right, did some more test and tried all possible camera settings to achieve the best results, not everything just in auto mode, but I really tried to get the best out of the camera.
Captured HDV footage in Ulead using 1080 50i profile. (I have demo version of PP2 so that don't work)
Rendered as DV footage.
Captured same footage as donconverted DV (on camera).

Split the screen in half, and inserted both bits, rendered to DVD...
I challenge anyone to tell me where the split is.

Needless to say, watching the hd footage, even converted to WMV using the fattest HDV profile, absolutley blows away any of the DV footage....THAT goes without saying. That is not the point however.

Next I took the HD footage and created a DVD without any other mods to the fotage. Once again I cannot tell any benefit to this route.

My final conlusion, unless someone can prove me wrong is the following.

HVR-A1 recorded in HDV and downsampled in camera delivers the same results as apposed to downconverting in software when final result will be a DVD.
Compare that footage to the same shot on the PD-170, the PD-170 is marginally sharper in every instance.

In (my) conclusion.. The HVR-A1 delivers great pictures when paired up to a 3 CCD DV camera such as the PD-170 or VX 2100 and the footage can easliy be intercut without any big hassle. Possibly +10% contrast and -10% brightness filters can be applied to closer match the footage of the other cameras. When used as a DV camera, the HVR-A1 delivers best results when shooting in HDV and downconverting (either onboard or in post) as apposed to shooting in DV modes from the outset.

These are just my findings over the last two weeks of testing, and my oppinions only. Others may get better results, but for me it's clear. The HVR-A1 is a great 'additional' camera.

Good luck all...

[Tom Hardwick]

It's really good to read your tests Mike. For years I shot weekly weddings with a pair of VX2ks. As far as image quality goes I could always better it with the DVX100A, but that camera was never as nice to use, so the Sony it was. Looking back at my work the pictures positively glow.

Then I changed to the Z1 & FX1, primarily because of the 16:9 aspect ratio and the far better camera ergonomics. I shot in DV mode because I am only producing normal DVDs for clients, but put simply, I cannot match the picture quality of the old VX, and I don't mean just in low light.

So rather than downconvert between chips and tape, I decided to shoot HDV and downconvert between tape and pc. The results are marginally better. I've not done controlled tests such as yours, but I shoot lots of weddings with the same kit, same style, same skill level. I'm very critical of my own work.

OK, I'm prepared to accept that (say) Canopus' Procoder will give me better downconversion, but I'm just sad to announce that the Z1 and FX1 simply cannot match the VX/PD for outright picture quality in SD mode. So all this talk of black stretch, picture processing, HD Zeiss lenses and T* multi-coating is worthless breath, because up there on the big screen the VX still looks better.

Of course I accept that the Z1 is an HDV camera that blows away the VX when it's allowed to, and that it's only backwards compatible (ie produces SD DV) as a favour to me. I'll just have to live with this, but I do so with a daily sigh of sadness.

tom.

[Mike Duncan]

Originally Posted by Tom Hardwick

It's really good to read your tests Mike. For years I shot weekly weddings with a pair of VX2ks. As far as image quality goes I could always better it with the DVX100A, but that camera was never as nice to use, so the Sony it was. Looking back at my work the pictures positively glow.

Then I changed to the Z1 & FX1, primarily because of the 16:9 aspect ratio and the far better camera ergonomics. I shot in DV mode because I am only producing normal DVDs for clients, but put simply, I cannot match the picture quality of the old VX, and I don't mean just in low light.

So rather than downconvert between chips and tape, I decided to shoot HDV and downconvert between tape and pc. The results are marginally better. I've not done controlled tests such as yours, but I shoot lots of weddings with the same kit, same style, same skill level. I'm very critical of my own work.

OK, I'm prepared to accept that (say) Canopus' Procoder will give me better downconversion, but I'm just sad to announce that the Z1 and FX1 simply cannot match the VX/PD for outright picture quality in SD mode. So all this talk of black stretch, picture processing, HD Zeiss lenses and T* multi-coating is worthless breath, because up there on the big screen the VX still looks better.

Of course I accept that the Z1 is an HDV camera that blows away the VX when it's allowed to, and that it's only backwards compatible (ie produces SD DV) as a favour to me. I'll just have to live with this, but I do so with a daily sigh of sadness.

tom.

Tom.
I concurr whole hartedly. Right here, right now, as I can deliver format to 95% of my clients, HDV is worthless (to me), so I use the HVR-A1 in the situation it is most usefull in. I.e. - an additional camera complimenting my others, but not as a main camera. I still love it's form and functionality and I'm glad I bought it. I even still use my Sony HCR-DC96 handycam as well for the 'Blairwitch' look and intercut all 4 camera's without a problem. I think it all comes down to the situation and uselfulness of an individual camera.

Personally I think my 74cm CRT t.v. has a better, brighter, more colourful picture than any new LCD or plasma screen, and my footage and music videos look very professional. So all this HDV 'hype' is great for a one off situation, but in the real world, it's useless to me....at this point.

[Mikko Lopponen]

[QUOTE=Mike Duncan;725991]

HVR-A1 recorded in HDV and downsampled in camera delivers the same results as apposed to downconverting in software when final result will be a DVD.

If you're doing interlaced then yes. But doing progressive from hdv in post will look so much better than doing progressive from sd that it's not even funny.

Possibly +10% contrast and -10% brightness filters can be applied to closer match the footage of the other cameras.

Mpeg2 decoding (hdv) is usual in the 16-235 luminance range and dv-encoding is 0-255. That means there will be a brightness difference if the player doesn't convert the mpeg's luminance to pc-values.

[Douglas Spotted Eagle]

Right, did some more test and tried all possible camera settings to achieve the best results, not everything just in auto mode, but I really tried to get the best out of the camera.
Captured HDV footage in Ulead using 1080 50i profile. (I have demo version of PP2 so that don't work)
Rendered as DV footage.
Captured same footage as donconverted DV (on camera).

Split the screen in half, and inserted both bits, rendered to DVD...
I challenge anyone to tell me where the split is.

You assume that Premiere has the same downsampling abilities as other applications. It does not. If you read through *most* of the posts on this subject, you'll realize that some apps don't do the same job as others. FCP for instance, was horrible for downconverting whereas Canopus Edius and Sony Vegas are not. Premiere (IMO) also isn't up to par. The in-camera downconvert will always be inferior for one primary reason; you're converting to 4:1:1 prior to transcoding to 4:2:0, whereas staying HDV through the workflow and converting in post not only is a 4:2:0 workflow throughout, but you also are presented with significantly more options at the end of the project that you simply do not have if you convert to DV from the camera.
Weekdays, I capture everything as HD. On weekends at my "hobby" work, I capture between 25 and 40 projects that have a length of 5 minutes that I'll never, ever again see, and don't care about. For these, I downconvert for capture, simply because my laptop used for editing these small pieces isn't up to par for fast HDV editing, plus the render to MPEG2 for DVD is much faster from a DV file vs the greater resolution being downconverted, transcoded.
Try your experiment again with Sony Vegas or Canopus Edius, and you'll see an end difference. Or, have a peek at one of the VASST HDV training discs, and there is a split screen there as well. The difference isn't monstrous, but it's not insignificant, either.

[Mike Duncan]

Hello Douglas.

Let me just then finalize my conclusions.

In my experience, and with the software at hand, I have drawn my conlusions. I.e. What can I do with my camera today, at this pint in time.

I of course have not tested every bit of software out there, and neither do I plan to. I am using what I have and need to establish findings according to what I have and can actually produce.

Thanks never the less for your input, I realize there are other methods out there.

[Douglas Spotted Eagle]

Originally Posted by Mike Duncan

Hello Douglas.

Let me just then finalize my conclusions.

In my experience, and with the software at hand, I have drawn my conlusions. I.e. What can I do with my camera today, at this pint in time.

I of course have not tested every bit of software out there, and neither do I plan to. I am using what I have and need to establish findings according to what I have and can actually produce.

Thanks never the less for your input, I realize there are other methods out there.

Fair enough, but it's important for others reading this thread to understand that your conclusions are drawn from a choice in tools that do not necessarily reflect all aspects of a standard.
Some editing applications for example, have horrible MPEG encoders as part of their package. Does this therefore mean that all NLE systems are terrible for encoding? Of course not.
I absolutely agree; if you're shooting HDV and aren't concerned about final output quality and/or output options down the road, *and* using Adobe Premiere or older versions of FCP, then you'll speed your workflow by downconverting in-camera as opposed to working with HDV on the timeline.

[Mike Duncan]

Hey Douglas, I couldn't find your split screens on your site..

[Douglas Spotted Eagle]

The split screen showing the differences in detail (both at normal view and at 3x view) are found on all the VASST HDV training DVDs (Canon A1/G1, Sony Z1/FX1, Sony V1/FX7). There is a low-quality version of it on YouTube under VASSTTraining as well. There are no screenshots/images of just the split screen on the VASST site.