GRO digital images

[Susan]

This morning I searched the GRO death index and found a new box on the results page. We can now buy certain images as a digital image for the cost of £2.50 and the image is available as soon as payment is complete. I decided it was time to buy some missing death certificates, but not all those I needed to buy were available as digital downloads. Experimenting with John Smiths who died in Wigan, the most recent death entries available as digital images were in 1887. Births for John Smiths in Wigan are available up to 1922.

This is the box which appears on the results page if an entry is available as a digital image -

 

[Susan]

I usually log in and go straight to the search page so I overlooked the first page after log in, which would have saved me a lot of trouble if I'd noticed it when I logged in!
Back in the newsletter of 30 April 2023, Peter speculated what the GRO intended to place in the gap. That question has now been answered.

 

[peter]

I'll be sending out a special newsletter shortly. I've been trialling Online View for almost 2 years, and I'm glad it has finally been opened up to all, as it has been difficult to keep such wonderful news to myself!

 

[LynSB]

I'll be sending out a special newsletter shortly. I've been trialling Online View for almost 2 years, and I'm glad it has finally been opened up to all, as it has been difficult to keep such wonderful news to myself!

Great news. Sad that I ordered a pdf a week ago but that’s life. I’ll certainly be making a lot of use of this new service.

 

[cfbandit]

I am so excited, I went through and ordered 3 images immediately. I was happy two of my educated guesses turned out to be correct. One however, her death remains elusive as the one I thought might be her, turned out to be an exciting story about an accidental drowning in Ratcliffe on the Wreake when my girl was in Leicester.

 

[KevinStortford]

Like the poster above I invariably go straight to the search and am so familiar with the relevant bits of the screens it didn't register the options box had expanded until I read the newsletter. I've just made my first use, which has confirmed from the informant, it is one of my 4xGt-Grandfather's brothers. The one thing I've found, which I've fed back, is the alignment is not quite right as per this screen clip extract. It's probably 13th December but there will be instances where guessing what is truncated is less easy. Do you come across anything like this in the testing Peter?

 

[peter]

Do you come across anything like this in the testing Peter?

Yes, it's a fairly common problem, and whilst I suggested a solution they were either unable or unwilling to implement it.

Generally you can figure out what it says, but if you're not satisfied you can ask for a refund. In the example about it could be either 13th or 30th (my guess would have been the latter), but you will know which is right from the date of registration.

 

[Kate]

Thanks for letting us know! I am very pleased with this and found my 2 x great grandmother's early death in 1879. Mostly it will be my 3 x great grandparents who are within the time frame, but it is affordable so I will definitely be tempted to spend more. Could be to the advantage of the GRO!

 

[Bob Spiers]

I tried out the GRO digital image on my Spiers GGF where I only had the birth year of 1840. It was more out of interest than anything else, which is as well since trying to decipher his date of birth is next to impossible. The birth was registered on the 18th March 1840 but I think I can just about make out 'February' in the 'When born' box; beyond that is just guesswork. I will likely let the matter rest at February.

 

[Stuart]

I tried out the GRO digital image on my Spiers GGF where I only had the birth year of 1840. It was more out of interest than anything else, which is as well since trying to decipher his date of birth is next to impossible. The birth was registered on the 18th March 1840 but I think I can just about make out 'February' in the 'When born' box; beyond that is just guesswork. I will likely let the matter rest at February.

I think I can make out "Twenty-sixth of February", though there's an element of imagination in that. One thing you miss by not having the top of the actual page is the sub-district name. That's still no substitute for the parish or village name, but it might help a bit. But the framing clearly needs to go a bit wider top and, especially, bottom; I suspect in a few cases more at the side is needed too.

 

[VTinOZ]

I tried out the GRO digital image on my Spiers GGF where I only had the birth year of 1840. It was more out of interest than anything else, which is as well since trying to decipher his date of birth is next to impossible. The birth was registered on the 18th March 1840 but I think I can just about make out 'February' in the 'When born' box; beyond that is just guesswork. I will likely let the matter rest at February.
View attachment 2356

Hi Bob, 1840 was a leap year, I think it says Twenty eighth day of February. No hope of making out the name of the place below though.
I had no difficulty purchasing images using my Australian Visa Debit card which was a relief. Peter may well be correct that the GRO's sales will be greatly enhanced by this new service.

 

[peter]

It was more out of interest than anything else, which is as well since trying to decipher his date of birth is next to impossible.

Just as well you didn't order a certificate - in the past I've had certificates that were just as difficult to read!

28th of February would also be my guess for the date of birth.

 

[KevinStortford]

Presume the parish is Harvington as per the 1861 & 1871 etc censuses?

 

[Helen7]

I think I can just about make out 'February' in the 'When born' box; beyond that is just guesswork.

Hard to read, certainly, but not impossible. To my eyes, it definitely says (something)-eighth of February 1840, so the twenty-eighth is the only option. And the birthplace looks to be 'Parish of Harvington' if indeed Harvington is the place mentioned in the Informant box.

 

[ChalfontR]

Like the poster above I invariably go straight to the search and am so familiar with the relevant bits of the screens it didn't register the options box had expanded until I read the newsletter. I've just made my first use, which has confirmed from the informant, it is one of my 4xGt-Grandfather's brothers. The one thing I've found, which I've fed back, is the alignment is not quite right as per this screen clip extract. It's probably 13th December but there will be instances where guessing what is truncated is less easy. Do you come across anything like this in the testing Peter? View attachment 2353

There were significant issues with alignment during the closed testing phase. I was one of the first testers, like Peter, and had a long discussion with their team about it. They explained that the GRO copy registers were scanned, many years ago now, page by page , with each scan being a whole page. When you order a digital image, the system accesses the relevant page image and then automatically crops it down to where the single entry you want "should" be. If the original page scan was misaligned in any way, and it seems many were, then the image you get will be slanted or have the top or bottom missing. I think about 10%-20% of the many images I've ordered since the original trial started have had alignment problems sufficient to make some of the information unreadable (or missing entirely).

The images are exactly the same as those reproduced on paper certificates and as pdf files, but the difference is that those have a human operator involved in the production process who will check and reposition the scan if necessary.

Perhaps GRO can implement some sort of "AI" driven solution, but I don't think the funding is there for that. Otherwise the only options are for a complete rescanning of the pages (again, no funding), or allowing the whole page to be viewed when a single record is ordered, which is what the (free) Irish record system does.

The good news is that if there is an alignment issue reporting it as an image problem to GRO gets a refund of your £2.50 within a couple of days ( or did during the closed trial anyway). I believe, but haven't had it confirmed, that when a problem is reported that particular record is removed from the available database, so the digital image ordering option will no longer be available.

 

[peter]

When you order a digital image, the system accesses the relevant page image and then automatically crops it down to where the single entry you want "should" be.

That's good news - I couldn't get a straight answer to that question, and I wondered whether they no longer had the scanned pages - the cost of hard drive storage was 100 times higher when the registers were scanned, and the DoVE project was way over budget.

Perhaps GRO can implement some sort of "AI" driven solution, but I don't think the funding is there for that. Otherwise the only options are for a complete rescanning of the pages (again, no funding), or allowing the whole page to be viewed when a single record is ordered, which is what the (free) Irish record system does.

My proposed solution was to widen the window so that in all but the worst cases you'd get the information you paid for. However this might require secondary legislation since it would reveal information from adjacent entries: note that in the FAQs they say "The way in which the original register page was captured digitally may be misaligned, which means that the image itself may contain partial details of another entry (although this will not identify the other entry). "

 

[Bob Spiers]

Presume the parish is Harvington as per the 1861 & 1871 etc censuses?

...And the birthplace looks to be 'Parish of Harvington' if indeed Harvington is the place mentioned in the Informant box.

Thanks to you both and actually I have pre-knowledge that my GGF (George) was born in Harrington (in the parish of Harvington) in Worcestershire and falling under Evesham, Worcestershire for census purposes.

 

[Bob Spiers]

To my eyes, it definitely says (something)-eighth of February 1840, so the twenty-eighth is the only option

28th of February would also be my guess for the date of birth.

I think I can make out "Twenty-sixth of February",

Hi Bob, 1840 was a leap year, I think it says Twenty eighth day of February.

Thanks to all and although I have tried hard to decipher what is written I have never got beyond February; and still cannot. I shall record in my Tribal Pages that the consensus viewpoint is the 28th February, but might just be the 26th.

Although his father is shown as a (Farm) 'Labourer' George was a Carpenter and Joiner by trade and later referred to himself as a Cabinet Maker. I have one of my earliest sepia type posed photographs of George taken in the 1880-1890's looking very (very) dignified with a wonderful moustache and goatee (type) beard, sporting a bow tie and stiff collar. Long ago captured in digital format from an original I believe came via a cousin.

 

[Stuart]

I agree it's more likely to the the twenty-eighth. I should have stopped to do what I've suggested before in such cases: look for a similar word. In this case the registration date provides examples of "th of" at the end of a line, which can be matched pretty well in the birth date. Even counting the squiggles to the left, if you don't think you can identify the letters, suggests "eigh".

 

[ChalfontR]

You could refer it back to GRO as an image quality issue, which would get your £2.50 returned. That is so poor quality it is the type of record that GRO would sometimes produce a certificate by re-typing the info.

If the date is important - you could get a certificate from the local regsiter office who hold the original document ( GRO don't)