750 as good as F900?

[Jason Chan]

Hi,

 

The rental house has suggested that I should take up the F750 instead of F900 as he mentioned that they are the same if I'm going to shoot on 25p.

 

I know the F900 is the camera of choice for a lot of features, but not much has been heard of the F750 compared to the 900.

 

Can anyone tell me what's the difference between them? and if there is a significant quality improvement using the F900?

 

Thanks for any feedback given.

[Clive North]

Hi,

 

The rental house has suggested that I should take up the F750 instead of F900 as he mentioned that they are the same if I'm going to shoot on 25p.

 

I know the F900 is the camera of choice for a lot of features, but not much has been heard of the F750 compared to the 900.

 

Can anyone tell me what's the difference between them? and if there is a significant quality improvement using the F900?

 

Thanks for any feedback given.

 

Hi Jason - I have a 750 here in the UK and the way I would look at it is the F900 is the better camera certainly for transfer to film for large screen projection (George Lucas seems to know what he is doing after all!)

 

The 750 is excellent for 25p for television and I can see know reason to go for the more expensive option in that case.

 

Hope this helps,

 

Clive North

[Jason Chan]

thanks for the feedback.

 

We intend to do a film transfer. So I guess F900 will have a significant quality improvement over the 750 for such purpose. Correct me if i'm wrong.

[Stephen Williams]

Hi,

 

Can anyone tell me what's the difference between them? and if there is a significant quality improvement using the F900?

 

Thanks for any feedback given.

 

 

Hi,

 

The F900 is far better than a 750. IMHO they are in a different league.

 

Stephen

[Jason Chan]

Thanks for the input.

Good thing I check with you guys here. Otherwise, I would have been "persuaded" to use the 750 by the rental house.

 

Thanks again!

[Guest MDO]

Thanks for the input.

Good thing I check with you guys here. Otherwise, I would have been "persuaded" to use the 750 by the rental house.

 

Thanks again!

 

Hi

 

If you recording in HDCAM mode ie, using the on-board VTR the differences between the two are marginal.

The f900 head is 12bit. The 750 is 10bit. But... BOTH are downconverted to 8bit to get onto the HDCAM.

Both are compressed at 4.4:1

The f900 has a more robust lens mount and slightly improved prism (as well as all the extra speeds).

By the time you have xferred to film, these benefits are all but lost.

However, if you are doing CGI you ideally need to record direct to an external no/low compression 4:4:4 VTR, not HDCAM. But then you would be using a 950.

 

MD

[Stephen Williams]

Hi

 

If you recording in HDCAM mode ie, using the on-board VTR the differences between the two are marginal.

The f900 head is 12bit. The 750 is 10bit. But... BOTH are downconverted to 8bit to get onto the HDCAM.

Both are compressed at 4.4:1

 

Hi,

 

I do not agree that the difference between a 750 and 900 is marginal.

 

HD Cam records 3:1:1 Then it is compressed @4.3:1.

 

Stephen

[Guest MDO]

Hi,

 

I do not agree that the difference between a 750 and 900 is marginal.

 

HD Cam records 3:1:1 Then it is compressed @4.3:1.

 

Stephen

 

Stephen, the compression figure of 4.4:1 is taken from Sony's own spec:

www.cinealta.com/product/acquisition/pdf/HDWF900_e.pdf

My point is that the small differences between camera head, once squeezed onto HDCAM fail to show up in a big way. It's wishful thinking to believe otherwise. Why bother to go to all the hassle of external 4:4:4 decks if HDCAM is so great? A couple of extra bits of quantization will not make a world of difference once re-quantized to the 8-bit HDCAM standard. ..add a 35mill transfer to the equation (as opposed to a decent HD projector) and anyone will tell you that you would be very hard pushed to tell the difference.

 

MD

[Stephen Williams]

Stephen, the compression figure of 4.4:1 is taken from Sony's own spec:

www.cinealta.com/product/acquisition/pdf/HDWF900_e.pdf

My point is that the small differences between camera head, once squeezed onto HDCAM fail to show up in a big way. It's wishful thinking to believe otherwise. Why bother to go to all the hassle of external 4:4:4 decks if HDCAM is so great? A couple of extra bits of quantization will not make a world of difference once re-quantized to the 8-bit HDCAM standard. ..add a 35mill transfer to the equation (as opposed to a decent HD projector) and anyone will tell you that you would be very hard pushed to tell the difference.

 

MD

 

Hi,

 

I think the camera head on the 900 is a very important issue! Shooting for a film out from a 750 will look far worse than from a 900. Have you tested the cameras next to each othe? Its Day and Night in my opinion!

Film transfered to HDcam on a Spirit looks better than the output from a 750 or a 900 in my opinion. If HDcam was the limiting factor that would not be the case!

 

Stephen

[Guest MDO]

Hi,

 

I think the camera head on the 900 is a very important issue! Shooting for a film out from a 750 will look far worse than from a 900. Have you tested the cameras next to each othe? Its Day and Night in my opinion!

Film transfered to HDcam on a Spirit looks better than the output from a 750 or a 900 in my opinion. If HDcam was the limiting factor that would not be the case!

 

Stephen

 

What aspects of the image look "far worse" on 750 film out?

HDCAM is most definately a limiting factor, especially where CGI is needed. Recording the native bit depth at 4:4:4 with no compression will give the absolute best images. HDCAM does not do this.

I don't get your film to video analogy..a 35mm neg has a much higher native resolution than HD. This will show up on an HDCAM transfer, thats why it looks better than the playback off an 750 or 900. HDCAM is not the lowest common denominator in this respect... big differences between formats (HD vs neg Film) will come through to some extent, small differences (10bit vs 12bit) will not (unless you are filming test charts!). The whole arguement is academic if the entire film to screen process is taken into account. This is where 35mm die-hards come unstuck...yes, 35mm has a higher native res. than HD, but by the time it reaches the big screen, you are back to an effective res. of 2k (plus grain, weave, dirt and scratches) . An HD projection of HDCAM on a Barco wins hands down.

The original question was about the technical differences between the two cameras. Fps choice aside (and that is a big aside), the differences do not amount to much when shooting real life, not test charts. I have seen both, digitally projected using real life footage and no-one could tell the difference. You may be able to spot the difference, switching between the two on monitor when looking at a locked-off test card, but does this amount to 'night and day' for reality? Most of the differences between the two are more to do with clever marketing than huge technical differences. Check out the respective brochures to see why....then read the smallprint in the spec.

 

MD

[Stephen Williams]

The original question was about the technical differences between the two cameras. Fps choice aside (and that is a big aside), the differences do not amount to much when shooting real life, not test charts. I have seen both, digitally projected using real life footage and no-one could tell the difference. You may be able to spot the difference, switching between the two on monitor when looking at a locked-off test card, but does this amount to 'night and day' for reality? Most of the differences between the two are more to do with clever marketing than huge technical differences. Check out the respective brochures to see why....then read the smallprint in the spec.

 

MD

 

Hi,

 

I have been very dissapointed with the 750 & 730, I feel they look like DigiBeta. Ok that was using HD zooms not Digiprimes.

 

Just my personal view!

 

Stephen

[Jason Chan]

Great information so far.

I'll be using the F900 with HD zooms only. I hope it doesn't look like Digibeta.

and btw MDO, that's the exact thing the rental house told me. That I should go for the 750 to save cost since i'm not doing any CGIs.

 

 

Hi,

 

I have been very dissapointed with the 750 & 730, I feel they look like DigiBeta. Ok that was using HD zooms not Digiprimes.

 

Just my personal view!

 

Stephen

Edited November 16, 2005 by Jason Chan

[Stephen Williams]

Great information so far.

I'll be using the F900 with HD zooms only.

 

Hi,

 

I don't think you will regret spending a bit more on origination.

I never heard anybody say its a shame we shot on 35mm, it would have looked better on.........

 

Stephen

[Guest MDO]

Great information so far.

I'll be using the F900 with HD zooms only. I hope it doesn't look like Digibeta.

and btw MDO, that's the exact thing the rental house told me. That I should go for the 750 to save cost since i'm not doing any CGIs.

 

Of course it won't look like digibeta, and neither would a 750. There is something seriously wrong if an HD format ends up looking like an SD format! I dont know how Stephen was viewing his 750 material, but there is no way a 750 looks like digibeta. Maybe if the detail settings are set wrong it will give a video look, thats all I can think would do it. I say again, check out the spec on these cameras.

[Stephen Williams]

Of course it won't look like digibeta, and neither would a 750. There is something seriously wrong if an HD format ends up looking like an SD format! I dont know how Stephen was viewing his 750 material, but there is no way a 750 looks like digibeta. Maybe if the detail settings are set wrong it will give a video look, thats all I can think would do it. I say again, check out the spec on these cameras.

 

Hi,

 

Shooting with the 970 DigiBeta today. I would say it looks better than the 730/50 range!

 

Stephen

[Filip Kovcin]

Hi,

 

Shooting with the 970 DigiBeta today. I would say it looks better than the 730/50 range!

 

Stephen

 

hello there, mailing from poland.

 

i am using both 750p and f900 cameras. in tv world i cannot say that there is visible difference between the two. if you are using both cameras with the same path from start to end.

the think where you CAN see "BETA" look is ofcourse when shooting in interlace and then downconvert to standard definition. then, in my personal opinion, it really looks like beta. which is probably normal due to interlace origin. BUT, of course, as you all know here when shooting in progressive mode (either f900 or 750p) camera, there is no way to look like beta.

 

the only difference i can see somehow exist - is when you look at 750p material on the big screen (after transfering to 35mm) and have really bright parts in your shot - window etc. together with normal exposure situation - in those situation (but, again this is maybe subjective opinion) i saw difference in treating bright parts. f900 had better response to those details in bright lights. but i did not tested them side by side. this is opinion i get when saw two different polish movies shot on our cameras but with somehow similar lighting conditions. (normally or dark lited interior with bright exterior in the same shot)

 

but

 

the problems starts when hd is DOWNCONWERTED to SD. it really depends how is downconveerted. there are many ways to do this and of course results are different. (f500, computer downconversion, jh3 player, etc.) and if you are asking me - wrong approach in this moment can result in bad impression. recently (3 days ago) i did that with f900 and downconwerted material looked in some way worse than PAL signal (imx). i believe that that is similar with 970 digibeta (thismodel 970 is NOT 790!!!!) which has progressive mode (imx too), of course.

 

my "theory" is that if you are working for TV - it's better to work with 970, imx or other progressive cameras - if you cannot achieve PROPER downconverting path. working with HD material eiter f900 or 750p if downconverted badly will in my opinion deliver not so impressive picture. the answer is maybe simple (i am not engeneer or tv electronic guru, just user) if there is no downconversion - signal is untouched. if donconverted it MAY add artifacts...

 

another aspect of "beta look" is ofcourse DOF issue, but this is another story.

 

just my thoughts.

 

thanks,

 

filip

Edited December 20, 2005 by filip kovcin

[Stephen Williams]

another aspect of "beta look" is ofcourse DOF issue, but this is another story.

 

just my thoughts.

 

thanks,

 

filip

 

Hi,

 

The DOF issue is identical with any 2/3 inch chip camera! Am I missing something?

 

Stephen

[Filip Kovcin]

Hi,

 

The DOF issue is identical with any 2/3 inch chip camera! Am I missing something?

 

Stephen

 

Stephen,

 

of course, i just forgot to mention that i was thinking of 35mm DOF comparing to 2/3" DOF.

 

 

note: jump over following part if ewerytnig is clear for you

 

====================

2/3" DOF "look" is 2/3" look, not 35mm DOF look...

 

i just wanted to say that 35mm camera will always have it's own (of course) film look DOF, but 2/3" cameras, even when progressive, even if HD will always have that 2/3" - "digital" DOF look. i mentioned that because sometimes is difficult to explain to i.e. client, that despite all quality HD brings it will see it's own world in it's own way - which is connected with 2/3" sensors. so if someone is expecting DOF of 35mm in all situations - when shooting with 2/3" cameras - he is of course rather wrong. you can mimic film DOF by choosing proper focal lenth etc, but as you all know this is not possible always, due to (still existing) laws of physics :)

=====================

 

thank you,

 

filip

Edited December 20, 2005 by filip kovcin

[Guest MDO]

Hi,

 

Shooting with the 970 DigiBeta today. I would say it looks better than the 730/50 range!

 

Stephen

 

Stephen

 

Are you seriously saying that SD 25P rushes off a 970 viewed on a studio SD monitor look better than non-downconverted HD 25P rushes off a 750 viewed on a studio HD monitor?

[Stephen Williams]

Stephen

 

Are you seriously saying that SD 25P rushes off a 970 viewed on a studio SD monitor look better than non-downconverted HD 25P rushes off a 750 viewed on a studio HD monitor?

 

 

Hi,

 

I would like to see a direct comparison with the 970 up converted to HD, and the 750 downconverted to SD, with the same lens on both cameras.

I am fairly sure for Blue screen the 970 results would be better for an SD finish.

 

Stephen

[Peter J DeCrescenzo]

I would like to see a direct comparison with the 970 up converted to HD, and the 750 downconverted to SD, with the same lens on both cameras. I am fairly sure for Blue screen the 970 results would be better for an SD finish.

Hi Stephen: I'd love to see the above comparison, too.

 

I'd also like to see a comparison of a DSR-450WSL in 24p mode with Sony's SDI output option card installed ($19,300 US total list w/o lens) and recorded "live" onto a Rosendahl bonsaiDrive with its SDI option (uncompressed 10-bit 4:2:2 HDD recorder/player @ $3,150 US list) against the new Sony 970 Digibeta and 750 HDCAM, again using the same HD lens.

 

I'm not saying the DSR-450WSL would be "better"; I'd just like to see how these would look, especially when chromakeying and compositing is involved. Keeping in mind the huge price difference between these cams.

 

Info about the bonsaiDrive is available here:

http://www.bonsaidrive.com/bonsai.html

[Stephen Williams]

Hi Stephen: I'd love to see the above comparison, too.

 

I'd also like to see a comparison of a DSR-450WSL in 24p mode with Sony's SDI output option card installed ($19,300 US total list w/o lens) and recorded "live" onto a Rosendahl bonsaiDrive with its SDI option (uncompressed 10-bit 4:2:2 HDD recorder/player @ $3,150 US list) against the new Sony 970 Digibeta and 750 HDCAM, again using the same HD lens.

 

Peter,

 

I think the 450 would hold up well!

 

Stephen

[Guest MDO]

Hi,

 

I would like to see a direct comparison with the 970 up converted to HD, and the 750 downconverted to SD, with the same lens on both cameras.

I am fairly sure for Blue screen the 970 results would be better for an SD finish.

 

Stephen

 

Yes, the 970 would be better, for any situation because no conversions are taking place. But what about your quote that the 970 looks better than the 750? How do you back this up? It is pointless to compare top of the range SD to down-converted HD. If your end format is only ever going to be 25fps SD then shoot SD with the best lenses, shooting HD then downconverting is a waste of time and money. Digibeta, 790 or 970 (4:2:2, low compression) is a superior record format to HDCAM (3:1:1, if I remember, very high compression). I dont know why Sony chose to go down the "xxCAM" format for HD rather than the superior DIGIBETA format. I wonder if the next step will indeed be "HDDIGIBETA" tho thats quite hard to say! I guess the SR1 is in fact HD-digibeta or something very similar.

[Stephen Williams]

How do you back this up?

 

It is pointless to compare top of the range SD to down-converted HD. If your end format is only ever going to be 25fps SD then shoot SD with the best lenses, shooting HD then downconverting is a waste of time and money.

 

Hi,

 

Just an opinion of what I saw with my eyes, nothing more!

 

Shooting on film and downconverting to SD yields better results than shooting on SD or HD. One would expect HD downconverted to SD to look better too.

 

I am Sure a F900/3 would give a better result downconverted SD than a 970

 

Stephen

[Guest MDO]

hello there, mailing from poland.

 

i am using both 750p and f900 cameras. in tv world i cannot say that there is visible difference between the two. if you are using both cameras with the same path from start to end.

the think where you CAN see "BETA" look is ofcourse when shooting in interlace and then downconvert to standard definition. then, in my personal opinion, it really looks like beta. which is probably normal due to interlace origin. BUT, of course, as you all know here when shooting in progressive mode (either f900 or 750p) camera, there is no way to look like beta.

 

the only difference i can see somehow exist - is when you look at 750p material on the big screen (after transfering to 35mm) and have really bright parts in your shot - window etc. together with normal exposure situation - in those situation (but, again this is maybe subjective opinion) i saw difference in treating bright parts. f900 had better response to those details in bright lights. but i did not tested them side by side. this is opinion i get when saw two different polish movies shot on our cameras but with somehow similar lighting conditions. (normally or dark lited interior with bright exterior in the same shot)

 

but

 

the problems starts when hd is DOWNCONWERTED to SD. it really depends how is downconveerted. there are many ways to do this and of course results are different. (f500, computer downconversion, jh3 player, etc.) and if you are asking me - wrong approach in this moment can result in bad impression. recently (3 days ago) i did that with f900 and downconwerted material looked in some way worse than PAL signal (imx). i believe that that is similar with 970 digibeta (thismodel 970 is NOT 790!!!!) which has progressive mode (imx too), of course.

 

my "theory" is that if you are working for TV - it's better to work with 970, imx or other progressive cameras - if you cannot achieve PROPER downconverting path. working with HD material eiter f900 or 750p if downconverted badly will in my opinion deliver not so impressive picture. the answer is maybe simple (i am not engeneer or tv electronic guru, just user) if there is no downconversion - signal is untouched. if donconverted it MAY add artifacts...

 

another aspect of "beta look" is ofcourse DOF issue, but this is another story.

 

just my thoughts.

 

thanks,

 

filip

 

 

Hi filip

 

I agree with your points about downconversion, although you seem to have confused 'record format' with "look". The record format has little do do with the "look". For example, 35mm films are played back off videotape formats for broadcast tranmission. In the TV domain, this has absolutely no effect on the look of the film (assuming no NTSC conversions) the fact that it is now videotape is irrelevant. In the same way, downconverted HD transferred to Beta does not "look like Beta". Therfore there is no such thing as the "BETA" look, the difference between beta and HDCAM is down to the fact that HDCAM records more lines.

What defines the "film look" more than any one thing is shooting progressivly. It has nothing to do with any particular record format, it is a function of the camera head, not the record format. Actually, all progressive camcorders record two interlaced fields! The important bit is how the image is captured in the camera head; either as two interlaced fields OR one progressive frame which is then split into two fields for interlaced recording, Psf. When you view progressively shot images on a monitor you are in fact looking at an interlaced image. Thats how all TV monitors work, SD or HD.

 

I'm afraid that for too long, ex-film camera people have confused Video record format with "video look".

"Video look" can be defined chiefly by Interlaced-captured images. To a lesser extent, over-electronically sharpened pictures, burn out and DoF can contribute to a "Video look". Things such as grain, picture weave and sparkle are faults of film that may or may not be desirable.

 

Moving on to contrast handling, the f900 and the 750 when correctly set up to optimise the full range of the CCDs will have the same contrast range. The f900 has a higher quantization bit level, but this will not effect the contrast range. It will, in theory, effect the colour resolution but only if you use an external low compression VTR that maintains the higher bit depth. Like it or not , the differences between the f900 and the 750 in CAM mode are marginal (fps choices aside). BTW, a correctly set up digibeta will match HD cameras in terms of contrast handling (around 11 stops).